Competition Policies in Emerging Economies

Lessons and Challenges from Central America and Mexico

This page intentionally left blank

Competition Policies in Emerging Economies

Lessons and Challenges from Central America and Mexico

Claudia Schatan • Eugenio Rivera
Editors

Project of Canada's International Development Research Center (IDRC) and the Economic Commission for Latin America and the Caribbean (ECLAC)

"Strengthening Competition in Latin American Isthmus: National Policies and Institutions, Regional Coordination and Participation in International Negotiations"

Editors
Claudia Schatan                                        Eugenio Rivera
ECLAC, Mexico City                               Director of the Area of Regulation
Mexico                                                     Foundation Chile 21
www.cepal.org.mx                                   Santiago
cepalmexico@cepal.org                           Chile

A copublication with the
International Development Research Centre
PO Box 8500
Ottawa, ON, Canada K1G 3H9
www.idrc.ca / info@idrc.ca
ISBN (e-book) 978-1-55250-401-7

ISBN: 978-0-387-78432-8 (Hardcover)                  e-ISBN: 978-0-387-78433-5
ISBN: 978-0-387-78434-2 (Softcover)
DOI: 10.1007/978-0-387-78433-5

Library of Congress Control Number: 2008925636

© United Nations 2008
All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science + Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden.
The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights.

Printed on acid-free paper

9 8 7 6 5 4 3 2 1

springer.com

Foreword

The challenges faced by market competition have been more widely recognized in the Latin American region over the last few years. There has been renewed interest in antitrust policies, in modernizing various regulations and achieving greater transparency in the way firms operate. The relevance this topic has acquired has grown precisely at a time when the concentration of wealth has deepened regionally and globally. The lack of appropriate pro-competition legal and institutional frameworks during the privatization process of large public enterprises in the 1980s and 1990s and a great number of ensuing mergers and acquisitions have made possible frequent anti-competitive practices, adversely affecting consumers and the competitiveness of producers.

In a number of public utility services essential to the economy, large privatized firms, formerly under public ownership, often act as monopolies. These practices have spread internationally. Economic liberalization and digitalization have made it easier to invest capital in foreign markets, but little has been done to curb abuse of market power in many developing countries where they operate.

This book addresses competition policies in Central America and Mexico, particularly in the banking and telecommunications sectors, in which market distortions have led to low levels of efficiency and competitiveness. In the cases of both of these sectors, access to credit and a modern telecommunications system is vital for the constant innovation and efficiency of their services. On the other hand, access to these services has become part of the population's basic well-being. The arrival of foreign direct investment (both regional and international) in the banking system and in the telecommunications sector has not produced improvement in quality or more competitive prices of these services in most of the countries studied. An effectively enforced competition policy can go a long way towards strengthening these sectors, among others, especially in small economies where this policy faces many obstacles set up by strong economic and political interests.

The experiences of the seven countries studied in this volume are a valuable point of reference for competition policy officials, who are either adapting their laws in order to strengthen and adjust them to their own realities, or else are enforcing newly enacted competition laws for the first time. The studies included in this book also provide a wide range of experiences in the difficult relationship that usually exists between competition policy officials and sector regulators and suggest ways of improving their cooperation.

This volume contains the most important results of the project "Strengthening Competition in the Central American Isthmus: National Policies and Institutions, Regional Coordination and Participation in International Negotiations", funded by International Development Research Center (IDRC) and implemented by Economic Commission for Latin America and the Caribbean (ECLAC). Within this project 18 national studies were carried out: three in each country of the Central American Isthmus covering general competition conditions, competition and regulation policy in the banking sector, and competition and regulation policy in the telecommunications sector. These documents, together with three similar studies on Mexico, nurtured three chapters of comparative analysis in this book. Another two chapters show a country case study of competition policy application (Costa Rica) and a sector case study – competition within the banking system – in Mexico. Finally, the conclusions of the book are presented in an international perspective, where the experiences of third developing countries are brought up and enrich the findings of the book.

The efforts to improve competition conditions in the region have increased, but there is a long way to go. We can expect knowledge and experience to develop much more in the coming years.

José Luis Machinea
Executive Secretary
ECLAC

Acknowledgements

Our sincere gratitude to IDRC for having financed the project "Strengthening Competition in the Central American Isthmus: National Policies and Institutions, Regional Coordination and Participation in International Negotiations" which was executed by ECLAC between 2004 and 2006. This book is the second one published by ECLAC on competition policy which has been funded by IDRC. The first one was Competition Conditions and Policies in the Central America and Caribbean Small Economies (written in Spanish), coordinated by Claudia Schatan and Marcos Avalos (published by Fondo de Cultura Económica and ECLAC in Mexico in 2006). This book reflects the findings of more than 25 documents produced in these endeavours.

We deeply appreciate the work by all the consultants who participated in the IDRC/CEPAL project and made a direct or indirect contribution to this book. Besides Marcos Avalos, Fausto Hernández, Adolfo Rodríguez and Pamela Sittenfeld, who appear as authors, we would like to thank the background papers written by Claudio Ansorena, Pedro Antonio Argumedo, Edgar Balsells, Simon Evenett, Marco Fernández, Ricardo González, Greivin Hernández, Mauricio Herrera, Maribel Macías, Judith Mariscal, Francisco Molina, Gustavo Paredes, Diego Petrecolla, Antonio Romero, Marlon Tábora, Carmen Urízar, Leiner Vargas and Marlon Yong.

We are sincerely grateful to Fernando Ramírez and Juan Luis Ordaz for their excellent technical support to produce the comparative Chapters 2, 3 and 4.

This volume could not have been written without the cooperation of the competition authorities of the Central American Isthmus and Mexico, as well as other specialists and public officers of the countries of this region. Among them, special recognition is given to Isaura Guillén, Executive Director of Costa Rican Antitrust Authority (Commission to Promote Competition); Celina Escolán, Superintendent of Competition in El Salvador; Edgar Reyes, Director of Competition, Ministry of Economics of Guatemala; Santiago Herrera, General Coordinator, National Program of Competitiveness UCP-FIDE, Honduras; Julio Bendaña, General Director of Competition, Ministry of Promotion, Industry and Trade, Nicaragua; Gustavo Paredes, former President and Commissioner of the Free Competition Consumer Affairs Commission in Panama; Óscar García, Head of Analysis and Market Survey, Authority for the Protection of Consumers and Defense of Competition in Panama; and, finally, Eduardo Pérez Motta, President of the Federal Competition Commission of México. All of them contributed to the discussion in workshops and seminars which enriched the documents that have provided the main inputs for this book.

We also acknowledge the very valuable contribution of Andres Rius and Susan Joekes from IDRC to the substantive part of the IDRC/CEPAL project, which helped conceive solid studies, as well as Taimoon Stewart, who has done deep reflection on competition in small developing economies, for her thoughtful comments. Observations from Sebastián Sáez and Verónica Silva, from the International Trade Division of ECLAC, Santiago, Chile, were also very helpful. Other organizations which provided support and made an important contribution to the discussion of the background papers for this book were the Central American Bank for Economic Integration in Honduras; the Economic Policy International Center for Sustainable Development, National University of Costa Rica; the Economic and Social Research Institute, Rafael Landívar University in Guatemala; and the Investment and Development for Exports Foundation in Honduras.

We also want to thank Rebeca Grynspan for her support to the IDRC/CEPAL project and the development of the studies contained in this volume. In the same way, we are indebted to our colleagues Fernando Cuevas for generously sharing his knowledge on competition policies; Hugo Enrique Sáez, who supported the Spanish edition of the book; and Pedro Cote, for his great efforts in the communications area, which helped us have important feedback on the papers and discussion on which this book is based. We also thank Laura Gutiérrez for her efficient administrative support throughout the project, which made it possible for many people to collaborate in the IDRC/ECLAC project and the production of this book. We also thank Nicholas M.A. Papworth for the translation from Spanish into English.

Contents

This page intentionally left blank

Contributors

Marcos Avalos Bracho, Ph.D. in Economics, Essex University, England. He also has postgraduate studies in education teaching techniques on economics regulation at the JFK School of Government, Harvard University. He has been Economic Affairs Officer at the Economic Commission for Latin America and the Caribbean (ECLAC) in Mexico, professor at the Instituto Tecnológico de Estudios de Monterrey-Campus Ciudad de Mexico (ITESM), Centro de Investigación y Docencia Económica (CIDE), Instituto Tecnológico Autónomo de México (ITAM). Currently, he is full-time profesor at the Centro de Alta Dirección en Economía y Negocios (CADEN) at the Universidad Anáhuac. Among his more recent publications are: Condiciones y políticas de competencia en economías pequeñas de Centroamérica y el Caribe; Claudia Schatan and Marcos Avalos, Editors, FCE and ECLAC, 2006, Regulación en aerolíneas en México; Marcos Avalos and Víctor Valdes, CIDAC, 2006, Política de competencia en México: las fusiones; Marcos Avalos and Camilo Vidal, CIDAC, 2007.

Adolfo Rodriguez Herrera, Ph.D. in Economics, Catholic University of Louvain, Belgium. He worked at the University of Louvain as researcher and visiting professor (1989–1994). He has held different executive positions in the public sector of Costa Rica, such as Superintendent of the Securities Commission for eight years and was also adviser to the Presidency of the Republic (1995–1998), and more recently to the President of the Social Security Institute and to the General Manager of the National Bank of Costa Rica. He has worked as a consultant for IDB, The World Bank, UNDP and ECLAC. He has recently finished all the coursework of a postgraduate degree in psychoanalytic theory at the University of Leon and the Institute of Neurosciences and Mental Health in Barcelona (Spain). Currently, he is a member of the Board of Directors at the Service Regulatory Authority of Costa Rica.

Claudia Schatan, MA in Economics, University of Cambridge, England. She is Head of the Trade and Industry Unit at the Economic Commission for Latin America and the Caribbean (ECLAC), United Nations, in México. She has been professor and researcher at the Centro de Investigación y Docencia Económica (CIDE) in Mexico, and visiting researcher of different universities in the United States (Harvard Institute of International Development (HIID), Harvard University, Center for United States-Mexican Studies, California University in San Diego, California, among others). Among her more recent publications are: Condiciones y políticas de competencia en economías pequeñas de Centroamérica y el Caribe; Claudia Schatan and Marcos Avalos, Editors, Fondo de Cultura Económica and ECLAC, 2006; El medio ambiente y la maquila en México: un problema ineludible; Jorge Carrillo and Claudia Schatan, Editors, Libros de la CEPAL No. 83, México D.F. 2005.

Pamela Sittenfeld, MSc in Regulation, with emphasis in Utilities Regulation and Competition Law from the London School of Economics and Political Science. She holds a Law degree from the Universidad de Costa Rica. She was the ex-director of Costa Rican Antitrust Authority (Commission to Promote Competition) and worked as a consultant for the Economic Commission for Latin America and the Caribbean (ECLAC) in Mexico. Currently she is a member of the National Council for the Supervision of the Financial System (CONASSIF), a member of the Board of Directors for the Regulatory Authority of Public Services of Costa Rica (ARESEP) and a consultant at Weinstok in Costa Rica.

Fausto Hernández Trillo, Ph.D. in Economics, The Ohio State University. He is a professor of economics at the Centro de Investigación y Docencia Económicas. He has been Budget Director in the Finance Ministry in Mexico, and a member of the Research Department at the Mexican Stock Exchange. He also is the author of three books and numerous articles in academic journals. His research interests include fiscal issues and financial markets. He has also been a consultant for the Mexican and Guatemalan governments and several international organizations such as the IMF, the World Bank and the IADB. Currently, he is also the Director of the prestigious academic journal El Trimestre Economico.

Eugenio Rivera, Ph.D. in Economics and Social Sciences, Free University of Berlin. He has been professor and researcher at the Centro de Investigación y Docencia Económica (CIDE) in Mexico, Coordinator of the IDRC/ECLAC Research Project on "Strengthening competition in the Central American Isthmus: National policies and Institutions, Regional Coordination and Participation in International Negotiations", Head of the Regulation Division of the Ministry of Economics in Chile, and Chairman of the Preventive Antimonopoly Commission. Currently he is participating in the creation of the Regulatory Commission of Public Works in Chile. Among his more recent publications are: Teorías de la Regulación en la Perspectiva de las Políticas Públicas; New trends in the Latin American telecommunications market: Telefonica & Telmex.

Chapter 1
Introduction

Claudia Schatan and Eugenio Rivera

The need for an efficient functioning of markets has been gaining progressive recognition in the developing countries as a mechanism to supply goods and services at reasonable prices for consumers, inputs at competitive prices for producers and for levelling the playing field for potential competing companies, including small and medium-sized ones.

Between the 1950s and the 1980s, the control of prices, interest rates and imports in Latin America guaranteed the development of the national industry, as well as the access of population and producers to essential services and to banking credits at preferential rates. This situation became impossible to sustain after the mid-1980s. The over-indebtedness, the enormous fiscal deficits and a growing lag in the competitiveness of the productive sector, among other factors, ended the aforementioned policies. Therefore, an effort to promote the good functioning of markets was, and continues to be, one of the routes that can help these developing economies to ensure a suitable behaviour of economic actors. However, that has not happened in a spontaneous way despite the macroeconomic policies of opening and deregulation. That is why the competition policy becomes indispensable to eliminate distortions in the markets.

This book focuses on the development and the challenges that the competition policy faces in Latin American countries, with a special interest in the small economies of Central America. Most of the latter countries have incorporated competition policies with a considerable lag in their governmental agendas. Thus, the main aim of this book is to analyse the market distortions in that region, the legal and institutional competition instruments governments rely on and those that could be developed to face such distortions.

The competition policy – understanding by that, mainly, the antitrust law and the competition agency in charge of applying it – has faced serious difficulties in the Central American Isthmus. There is a fertile land for anti-competitive practices in this region given its numerous national markets of reduced size and its concentrated productive structure, especially in the non-traded product areas. These markets often shelter horizontal monopolistic practices (e.g. agreements among producers of goods or similar services to fix prices, to fragment markets geographically), as well as vertical practices (e.g. conditioning of input purchasing or input sales; establishing contracts of exclusivity in the distribution and sale of goods). It is particularly important for the competition agencies to distinguish concentrations that respond to a necessity of a greater firm size required by a technological innovation which will improve industry efficiency from the practices that search the increase in profit margins through monopolistic power.1

The scarce resources of competition agencies in countries with little human and financial capital call for a special effort to adapt the competition policies commonly used in developed countries to their own realities. Social conditions lagging behind, the great number of small and medium-sized firms and the exposure of competition authorities to be "captured" by groups of interest demand a careful design of competition policies in those countries (see Schatan and Avalos, 2006).2

Besides making a comparative and critical analysis of the Central American countries' experience (Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua and Panama), additional reference points are taken into account in this book, especially from Mexico. Among the Latin American countries, Mexico is the closest to Central America, not only geographically but also in competition policy subjects. In fact, Mexico has had a strong influence on the laws of this region, so that it is a forced reference when studying the countries of the Isthmus. Although the latter nation differs in size from the former ones, they share a similar cultural, legal and institutional background. The competition authorities in Mexico have had to find ingenious solutions, not always successfully, to solve difficult cases in a relatively weak institutional framework. A helpful aspect of the Mexican competition policy is the cumulative experience regarding the notification to the competition authority obligation prior to the merger.

It is important to point out that the combination of supports and resistances faced in order to engender the competition policy explains the degree of speed upon which such policies were introduced in the studied countries. Between 2004 and 2006, progress has been fast, since three countries enacted their competition law for the first time and two others revised them. The sequence of law enactment in the seven countries under study is the following: Mexico (1992, and amended in 2006), Costa Rica (1995), Panama (1996, and amended in 2006), El Salvador (2004), Honduras (February 2006) and Nicaragua (October 2006).3 Guatemala was still in the process of discussing a law project at the end of 2006. The experience gathered by the first countries that enacted competition laws has been very useful for the rest of the countries.

This book also analyses the specific conditions and policies of competition in two strategic sectors – telecommunications and banking – in six Central American countries and Mexico. These two sectors are crucial as much for the possibilities of

1 For greater details on restrictive competition practices, see UNCTAD (2004).

2 On the economic groups formation and the role played by them in Central America, see Segovia (2006).

3 The studies contained in this book were concluded in September 2006, i.e. before the Nicaraguan competition law enactment (October 2006); hence, this legal change is not taken into account in the chapters of this book.

advancing economies more quickly in the technological innovation and competitiveness as for increasing population's well-being. This is important since intercommunication has become a basic need for society, as has the access to credit and banking services. The Central American region, to a greater or lesser degree, still suffers from a lag in the access to these services. This situation is, partly, a consequence of an incapacity of oligopolies and monopolies (until recently public) to respond to the increasing demand of population and companies to these services. These two strategic sectors have also been highly regulated. However, this regulation has neglected the consequences of maintaining a highly concentrated market and has not demanded a competitive behaviour from companies.

In the telephony sector, the countries under study show different trajectories in privatization and liberalization, as well as in their market performance. In all of these countries, however, this service has improved substantially since the incorporation of the mobile telephony. This technology has allowed the expansion of telephone connections at an accelerated rate and has been able to respond to a long-time dissatisfied demand. Meanwhile, in spite of its expansion, the fixed telephony shows that it is still lagging behind. As will be seen in this book, the behaviour of agents in the market has shown different results in different market segments (mobile telephony and fixed telephony). Private monopolies that emerged from the privatization have had comparatively a much greater capacity to remain as the only company in the fixed telephony market (specially those companies to which the government gave an exclusive concession for several years) as compared to the mobile telephony sector, where there has been more space for the entrance of new companies. In any case, the companies that managed to hold on as monopolies at a national level have tended to dominate not only national markets but also regional and international ones. From this reinforced position, and helped by the convergence of networks and services, these monopolies also have progressively been positioning themselves in the market of mobile telephony, while increasing the competition in this market. Nevertheless, the increasing presence of regional duopolies in telecommunications and the competition policy's institutional weakness may frustrate the possibility to improve competition further. In a context of high barriers to the entrance and weak competition policies, the strong and generalized dominion of companies such as Telefónica and Telmex-América Móvil may encourage large companies to distribute markets among themselves and hence dominate most of them in the telecommunication sector.

Like in other countries and regions, the banking sector is one of the most State-regulated markets. This is because its inadequate operation can cause potential economic imbalances. In fact, this sector's lack of solidness led to diverse insolvency crises, especially during the first half of the 1990s. Since then, the countries of the region have subscribed the Basel Protocol, which has led to the introduction of international norms of financial supervision and to other mechanisms in order to guarantee the sector's stability. These measures have had the aim of improving the capitalization and financial soundness of the banking system. On the other hand, this has led to a greater market concentration. Nevertheless, the mere concentration in the market structure is not an indicator of abuse of market power. In fact, the information technology, the offshore operations and the greater facility for the free flow of international capitals have introduced a greater contestability in the banking system of Central America and Mexico. But important anti-competitive practices persist in this sector, and the Mexican case, studied in this book, is quite illustrative of this behaviour.

Costa Rica has been the country that accounts for the richest experience in the field of competition policy in Central America. This country not only was one of the first countries in signing an antitrust law and in creating a competition authority (Comisión de la Promoción de la Competencia, CPC), but its legal and institutional design also made the processing of hundreds of cases possible throughout ten years or more. This experience contrasts with the case of Panama, which has been limited by its unavoidable bond with the judicial system. The study on Costa Rica provides a detailed analysis of the most important cases taken by the CPC, from its creation, obstacles, the useful instruments available for their solution, to the learned lessons and the best international practices.

Although it goes beyond the objective of this book, it is important to point out that the obstacles faced by small economies to have an effective legal and institutional competition framework could be lessened through regional and international cooperation. Such cooperation has already begun to take place in trade agreements both among Latin American countries and with third countries. Nevertheless, a solid competition policy is needed at a national level. The competition clauses can include a wide spectrum, from cooperation involving consultations between competition authorities to the joint resolution of anti-competitive practices affecting countries in the region. On these lines, an important case is the agreement reached by the Andean Community, which has a Competition Commission with supranational powers. Countries in the region lacking an antitrust law of their own can use such an agreement.4 Nevertheless, the latter regional administrative entity has not reached the expected results, which has led to a reform in the agreement in 2005. Among the most influential changes carried out by the Andean Community are the greater competencies given to national agencies of competition. Therefore, strong national institutions are needed to support the growing mutual support in competition issues in a context of economic integration (Silva and Alvarez, 2006). Likewise, cooperation between developed and developing countries – key to approach anti-competitive practices of multinational companies – requires a greater institutional soundness from the latter countries. Otherwise, the competition policy institutions in the developed countries would prevent developing countries from the access to their information because they would fear that this could be handled in an inadequate way (Stewart, 2005).

The three following chapters have been elaborated based on seven national studies (six in Central America and Mexico). Chapter 2 analyses the competition policy laws in these seven countries. Chapter 3 presents a comparative analysis in the telecommunications sector for the same group of nations. There is also a

4 Bolivia and Ecuador.

regional comparative analysis for the banking sector in Chapter 4. Chapter 5 reviews the main case studies in Costa Rica, the country with the richest experience in competition policies in Central America. It emphasizes the obstacles faced as well as the lessons learned by the competition agency in that country. Chapter 6 contains a study on competition policy in the Mexican banking system which, notwithstanding its greater openness to foreign investment over the last years, is still considerably inefficient. This is a valuable point of reference for the Central American region where foreign capital is quickly starting to make its way into the financial market. Finally, Chapter 7 summarizes and interprets the main findings for Central America and for Mexico of the research conducted for this volume and relates those findings to the experiences of nascent competition agencies in other developing countries.

Bibliography

Schatan, C., Avalos, M. (2006), Condiciones y políticas de competencia. Economías pequeñas de Centroamérica y el Caribe, FCE and ECLAC, Mexico.

UNCTAD (2004), Model Law on Competition. http://www.r0.unctad.org/en/subsites/cpolicy/docs/Modelaw04.pdf.

Segovia, A. (2006), Integratión real y grupos centroamericanos de poder económico. Implicaciones para la democracia y desarrollo regional, ECA: Estudios Centroamericanos, ISSN 0014-1445, N°. 691–692, pp. 517–582.

Silva, V., Alvarez, A.M. (2006), Cooperación en políticas de competencia y acuerdos comerciales de América Latina y El Caribe: desarrollo y perspectivas, serie Comercio Internacional, N° 73, CEPAL, Santiago, Chile.

Stewart, T. (2005), Special cooperation provisions on competition law and policy: the case of small economies, In: P Brusick et al (eds.), Competition provisions in regional trade agreements: How to assure development gains UNCTAD, New York, Geneva.

This page intentionally left blank

Chapter 2
Markets in Central America and Mexico:
What Is Happening with Competition?

Eugenio Rivera and Claudia Schatan

Introduction1

Given their lack of competition, market functioning in Central America2 has become a topic of growing importance and attention. In recent years competition policy has evolved rapidly throughout Latin America, and has been included in the policy agenda of governments throughout the region.3 The purpose of this text is to evaluate the means by which some Central American countries have ratified competition laws, to analyse the characteristics of those laws or of the bills that are currently the subject of congressional debate and to identify the principal problems they experience while exploring possible solutions. Some existing laws or bills in Central America were influenced by those of Mexico, a country that has been applying such legislation for more than a decade (since 1992). That precedent justifies the inclusion of an analysis of that country as a reference point for much of the research carried out in this book.

This work assesses the reasons why those countries that have had such laws in place for a decade or more (Costa Rica, Panama and Mexico) have had to make changes to their competition laws in order to strengthen competition agencies. In the case of those nations that have only recently approved this legal and institutional framework, which is to say El Salvador (2004) and Honduras (2006), this study depicts the challenges they have faced in overcoming deep-seated opposition to approving such laws, as well as the extent to which they contain advances compared to their predecessors. Lastly, we point out the obstacles to approving such legislation in the two countries that have yet to adopt a competition law, Guatemala and Nicaragua. As was the case in El Salvador and Honduras, the

1 This chapter was concluded in September 2006. Nicaraguan competition law was enacted in October, 2006, so it is only considered as a law project in the analysis of this chapter.

2 For the purpose of this book, the Central American Isthumus includes the following countries: Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua and Panama. Belize is not included because not enough information on competition was available in that country.

3 Competition law is a legal instrument for broad application that is aimed at promoting and protecting market competition.

resistance in these last two countries emerged from an optimistic view of the market's self-regulating capabilities as well as a deep-seated opposition from within the business world. With the exception of Guatemala, however, a conviction developed throughout the region that the lack of a competition policy constitutes an insurmountable obstacle to achieving proper resource allocation and efficiency at the level of production. Within this analytical context, this chapter reflects on the experience and the most adequate competition framework for small economies and developing ones such as those of Central America. This reflection also seeks mechanisms for strengthening legal bodies that attend to competition problems, as well as the forms that help to politically value the issue of competition in a way that allows it to become a true priority in governmental agendas.

In synthesis, competition has recently assumed great importance and become the subject of considerable activity in the countries under study. Whether at a swift or plodding pace, countries have elaborated proposals, approved laws or sought to present their legislatures with major reform bills in an effort to improve competition conditions. This process has taken place at a time when considerable progress has been achieved on the level of competition policy worldwide. By 2006, there were 14 countries in the Americas with competition laws and a competition authority,4 approximately half of which had been adopted sometime around the middle of the 1990s. All other developing countries have since made progress on this front. Nevertheless, this remains a controversial subject for such economies as was apparent in 2001, when it proved impossible to include the Singapore issues, including that of competition, in the Doha Round of the World Trade Organization (WTO). Although Doha took up the question of competition (in response to a proposal from the European Union), it did so in the specific hope of keeping the benefits of trade opening from falling under the control of major corporations that might ally in cartels or adopt other anti-competitive practices to assure themselves of monopoly-scale profit margins. Many developing countries have expressed the need to expand their ability to implement competition policies by assuming multilateral commitments under which they could apply trade sanctions whenever competition rules are not respected.

There exists an undeniable need for competition policy in Central American countries.5 As with most developing countries, during the decade of the 1980s the region experienced a foreign-debt crisis and a dramatic upswing in oil prices that led to a scaling back of the economic intervention capabilities of governments, while leaving the market to assume a much larger role in the resource allocation within the economy. The deep distortions markets experienced following the privatization of large-scale publicly owned companies and the deregulation of prices and trade, to name a couple of key developments that got under way in the

4 International Competition Network (ICN) web site [online] (www.internationalcompetitionnetwork.org/).

5 While many types of economic policies affect the conditions of competition, such as trade policy, we will not focus on such matters because they are beyond the scope of this study.

mid-1980s, show that such markets have failed to spontaneously operate in a healthy manner, thereby making it necessary to regulate the behaviour of industry protagonists. Owing to a series of conflicts of interest between various groups within each country, several governments failed to intervene as needed.

The circumstances described above, their impact on economic growth, as well as their adverse effect on distribution and competitiveness dispelled optimism as to markets' self-regulating capabilities. The result was to make competition policy an indispensable aspect of a successful economic reform.

In what follows, Section I develops a conceptual framework for facilitating the analysis of competition policy in small developing economies (SDC). Section II studies the processes that led to ratification of competition law in Costa Rica, El Salvador, Honduras and Panama, as well as the obstacles to approving such legislation in the remaining countries in the region, including those that only recently managed to get such laws passed. We also analyse the factors that are weighing either in favour of, or against approval of, such bills that are currently under discussion. Section III offers a detailed analysis of the characteristics of the legal frameworks that are already in place or which are currently being considered. This section also takes up the institutional characteristics of the defence of competition in Mexico, Costa Rica, El Salvador, Honduras and Panama as well as the institutional environment envisioned in bills that are pending approval by some countries in the region. Section IV considers the role of the judicial system in applying such laws while exploring the reasons underpinning the grave problems encountered in trying to achieve an effective defence of competition. In Section V, we explore competition conditions prevailing in some markets in the region, as well as some pertinent cases that have been resolved in Costa Rica and Panama. This chapter ends with a series of conclusions in Section VI.

I. Competition Policy: A Necessity in Small and Developing Economies?

1. General Context

Two decades after most countries in the region introduced a series of economic reforms, several of which were directed towards improving market functioning in order to optimize allocated and productive efficiency, the results have generally proven to be frustrating on the level of competition conditions, although some signs are encouraging. The persistence of barriers to entry, especially those related to the ability of firms and business associations to influence public policy, leads to a variety of distortions that affect the development of market competition and which have proven impervious to change despite trade openings and other pro-free trade economic reforms.

At the outset of the economic reform, doubts emerged as to the need for, or usefulness of, including competition laws as part of the reform. Powerfully influenced by analytical focuses such as unrestricted market access theory (Baumol et al., 1982), some analysts maintained that to the extent that tariff and non-tariff trade barriers, as well as legal entry barriers to various industries, could be swept away by an active deregulation process, markets would function better and even optimally. Some writers holding a different perspective questioned the convenience or possibility of applying a competition policy in a context marked by incompetent governmental institutions or, worse yet, that could be dominated by private interests. From these analysts' perspective, a maximalist application of competition law could inhibit corporate freedom and restrict the rewards that firms should otherwise derive from the functioning of markets (Khemani and Dutz, 1996).

International experience has been categorical in showing that even when the liberalization of international trade introduces important competitive pressures to domestic markets, it offers no direct guarantee that these will function properly. In fact, while products more easily enter markets, there is nothing to keep importers and distributors from taking unfair advantage of their market power when marketing their products. In this way trade liberalization in and of itself does nothing to eliminate the propensity of firms to adopt anti-competitive practices. Similarly, privatization in no way assures that companies will behave better than their public sector predecessors as they are capable of erecting private anti-competitive barriers in the place of state ones (Cernat and Holmes, 2004; Schatan and Avalos, 2006).

It is worth noting that even when large and developed countries enjoy conditions propitious for market competition, some argue the existence of innate imperfections derived, for example, from the existence of change costs, especially for non-standardized products or commodities such as sugar, flour or other goods with characteristics that are difficult to differentiate. As Klemperer (1995) has pointed out, the cost of change arises out of the need for compatibility between newly acquired items and existing equipment in order to avoid the transaction costs that arise from having to change the supplier. Other costs that derive from change are learning how to use new brands, those related to uncertainties as to the quality of unfamiliar brands, the loss of discount coupons and similar mechanisms and, lastly, the psychological costs that accompany change or the abandoning of brand loyalty for non-economic reasons. All of these costs tend to raise prices and to provoke oligopoly-related deadweight losses, thereby discouraging the entry of new companies and reducing market competition.6

6 From a different perspective, some analysts substantiate the need for competition legislation on the basis of three factors. First, there is no proper controversy resolution mechanism owing to a lack of the litigation legislation needed to assure a perfect adjudication. Second, the costs for obtaining information or the transaction costs that may be incurred by those affected by a monopoly practice prove to be greater than the costs of those who engage in such practices and, third, it is less costly to adjudicate through a specialized body than by a regular civil court (Sánchez Ugarte, 2004).

The anomalies described above tend to occur in economies that enjoy competition-conducive conditions, which is to say large and developed countries, but the irregularities are much greater in the markets of developing economies and, even more so, in those which are small.

2. Small Economies: An Elusive Concept

Proposals that tend to formulate and win approval for competition promotion legislation have sparked a series of questions and debates among SDC. These include: (i) whether such economies need competition legislation or, in contrast, whether trade liberalization and deregulation suffice to generate a self-regulating market economy; (ii) what SDC means when speaking of market competition; (iii) whether the specific characteristics of such countries demand particular types of competition laws and policies or require an adaptation of the existing ones in the way in which they should be applied; and (iv) in light of the characteristics of such nations should competition policy add to its list of goals helping to achieve social objectives?

This study focuses primarily on Central American countries, which are small and developing economies. There is a limited body of literature analysing small economies and competition, and even less dealing with the topic in relation to SDC. Further complicating matters, the authors of such literature employ different concepts of what constitutes a small economy, and some even regard this as an unnecessary analytical category. In our view, the Central American economies are characterized by specific features that demand that the competition policies applied in large developed economies – which are by no means homogeneous – undergo a certain degree of adaptation.

In practice there is no specific view of competition for SDC. An initial focus for dealing with small economies that tend to be better developed was provided by Gal (2001), who defines a small economy as one that in the case of most industries is capable of sustaining only a reduced number of competitors in the face of limited domestic demand. The factors that help determine the scope of the market include the size of the population, how it is distributed and trade openness. According to this logic, a reduced market affects the three principal indicators of social well-being: efficiency in resource allocation, efficiency in production and dynamic efficiency. According to this outlook, three characteristics of small economies make competition in their markets more complicated: high degrees of concentration, diverse entry barriers and suboptimum production levels, which is to say below the minimum efficient scale (MES) that imposes upon them elevated production costs (Gal, 2003). The principal consequence of these considerations for competition policy in small economies is that they must make economic efficiency their principal objective because they are in no position to sacrifice it in favour of achieving broader objectives such as social goals. In this regard, Gal proposes that small economies, unlike the large ones, approve concentrations that expand the size of corporations so as to broaden dynamic efficiency, even as such transactions tend to curb competition. In this regard, De León (2006) points out how difficult it is for a competition agency in a small developing country to determine up to which point the actions of companies that expand their size favour dynamic efficiency, an issue tightly linked to technological innovation and to what extent actions such as mergers and acquisitions undermine competitors.

Gal has her critics, including Simon Evenett (n.d.), who believe that Gal's concept of small economy fails to offer characteristics that truly distinguish them from the largest economies. The alleged peculiarity that these countries display on the level of firm size (MES) applies to very specific cases in which imposing barriers to entry exist. From our point of view, Gal's analysis is mainly based on the situation of countries that are highly protected and to a certain extent isolated from the international economy in ways many countries experienced before the economic reforms that began in the 1980s became generalized. Nevertheless, her analysis remains pertinent for markets of non-tradable services.

On the other hand, Gal's concept of small economies (2001) is too skewed if we recall that this category extends to include Australia, Canada, Israel and New Zealand. In such a context it is difficult to analyse through the prism of competition conditions the sort of SDC that exist in Central America, especially the smallest of these such as the Caribbean nations studied by Stewart (2006).

It is interesting to note the opposite side of small economies from the one that Gal contemplates, which is to say those of Caribbean countries that Stewart (2006) studies. In reality each of these two groups suffers very distinct market problems. Stewart argues that for many countries of the Caribbean Community (Caricom), the problem resides in the extent to which such small economies make it difficult for competitive firms to exist, thereby limiting their capacity for technological progress. Rather than generate their own technology, countries on these islands import it, so they invest practically nothing in their own research and development, thereby limiting their ability to generate dynamic efficiency. She adds that there are "sub-optimal levels of production, since a considerably large fraction of all output is produced in sub-optimal volumes and sub-optimal plants, leading firms to be inefficient and internationally uncompetitive"7 (Stewart, 2006).

Furthermore, Stewart holds that the insertion of small economies into the global economy and international trade exposes them to competition from multinational corporations, which are capable of displacing local companies and, in the process, gravely undermining employment and the provisioning of food in such economies. According to the author, in such instances it would be totally justifiable if the objectives of competition policy were to be subordinated to social concerns.

We may deduce from this that the problem that the tiny countries of the Caribbean encounter from trade opening is the substitution of the domestic production of basic goods by international sources and its grave impact on employment

7 Quoted from the author's original text in English (unpublished in such language), p. 5.

with no possibility of generating additional economic activities, thus Stewart's conclusion on the need to subordinate competition policy to social objectives.

Central American nations stand somewhere between the two extremes studied by Gal (2001, 2003) and Stewart (2006). Trade liberalization has not produced a displacement of local production in these countries on the scale that Stewart describes. Within such nations there exists a wide range of products that can be produced by companies in the category of medium-sized enterprises (MSE), or the same effect can be achieved through exports to the rest of the region or countries beyond.

For SDC on the scale of Central American countries, the elimination of entry barriers to new products may expand economic activity. There are greater similarities between Central American and Caribbean countries in the extent to which economic and political power is concentrated into a few hands, a consolidation which facilitates collusion and other anti-competitive practices. In relation to such features, Gal (2004) emphasizes these societies' lack of a competitive culture, the political influences exerted on competition authorities and the interests of groups linked to major corporations among others characteristics8 that greatly limit the possibility of applying an effective competition policy. In such a panorama, Gal suggests making a priority of competition promotion and competition advocacy9 as a means for discouraging anticompetitive practices and postponing the formal establishment of legal and institutional competition channels. From another perspective, however, she advises that a competition authority that monitors and acts in cases of anti-competitive practices may also serve as a means of competition advocacy combined with significant effects designed to dissuade such behaviour.

In practice, economic reforms in developing countries have been – or are in the process of being – accompanied by the creation of indispensable regulatory and competition institutions. Nevertheless, these bodies at first tend to be characterized as congenitally weak and are suspected of tending towards an incorrigible tendency towards arbitrariness. These bodies' weakness is accentuated when they are dependent on judicial-deficient mechanisms to review their rulings, frequently neutralizing their effectiveness.10 In regulatory spheres, this situation has frequently translated into excessively generous tariff policies that at least partially explain the possibility for incumbents to expand their companies more than they would have been able to under genuine competition conditions, for example by means of an

8 It is noteworthy that the author points out that in developing countries the intention of achieving economic efficiency and competitiveness cannot be isolated goals and that competition must at times be subordinated to certain social objectives, especially those which concern distributive matters. These contrasts with the outlook this writer expressed in earlier works: 'Therefore, social objectives should have little or no weight in the formulation of competition policies in small economies' (Gal, 2001, p. 1452).

9 Competition advocacy generally refers to two types of activity: those that consist of promoting competition culture and the issuing of opinions on the impact of competition laws and norms.

10 This situation contrasts with that of developed countries. In England, for example, jurisdictional authorities generally assume that the regulatory authority serves the public interest and, as a consequence, firms are reluctant to carry their disputes before such judicial authorities.

aggressive policy of takeovers.11 In the specific field of competition, the panorama described above has contributed to undercutting the effectiveness of competition policy that in some countries has led to a sense of generalized impunity in this field. A case in point is the extent to which such violators excessively recur to an amparo (in the region's legal systems, an amparo is a frequently used appeal alleging a violation of constitutional rights that can act as a stay or injunction against government decisions, in both regulatory and judicial proceedings). In this way institutional deficiencies, particularly on the level of competition, can obstruct the development process by deterring economic agents from adopting salutary practices. Notwithstanding these problems, as we shall see further in this text, the countries under study have made progress towards overcoming such problems. But while the Central American countries, as well as Mexico, remain far from achieving a truly effective competition policy, it is encouraging to see the extent to which they have strengthened the process both as they have managed to draw on the experience accumulated over at least a decade into greater efficiency and in the extent to which the judicial system has grown more skilful (Costa Rica is a good example of this).

3. Some Additional Peculiarities of Small Developing Economies

Small and developing economies are highly susceptible to anti-competitive practices due to a series of issues that we dealt with in Section I of this chapter. Under these circumstances, some policy instruments assume particular importance, especially regulation. In effect, whenever several markets traditionally characterized by the presence of natural monopolies have been progressively undercut by competition the possibility has been posed of eliminating that sector's regulatory authorities and leaving supervision of such industries exclusively in the hands of competition agencies. New Zealand is the paradigmatic example in this regard. Nevertheless, it has become clear that over an extended period of time aspects peculiar to natural monopolies coexist with those of a competitive environment, thereby assuring the need to maintain regulatory policies. In effect, it is necessary to sustain tariffication in many of these industries and anti-competitive behaviour acquires forms that demand the intervention of agencies that specialize in the sector. Such situations assume even greater significance in the countries that we analysed and which face international operators in these types of sectors.12

In addition, developing countries and, even more so, the small ones in this category are in need of regulations that afford foreign investors security. In the case

11 For the case of telecommunications, see Mariscal and Rivera (2005a, b, 2007).

12 In this regard see the chapters of this book that analyse competition in banking and telecommunications.

of the countries we deal with in this study, Levy and Spiller (1996) offer an analysis of the advisability of defending investors against the risk of expropriation that should be taken into account. The regulatory framework should stimulate private investors, especially those from abroad, by employing two complementary mechanisms to restrict arbitrary administrative actions: first, by imposing substantial restrictions on discretional administrative action and, second, by putting into place formal and informal restrictions on making changes to regulatory systems and institutions that are designed to reinforce such restrictions. In the countries under study, the job of guaranteeing the rights of foreign investors is further backed by investment accords between countries. Such agreements are designed to provide foreign investors protection and non-discriminatory treatment, as well as to establish clear rules on earnings repatriation, controversy resolution mechanisms and the rules under which compensation is to be determined in the case of an expropriation, to list a few. While these concerns may appear reasonable in principle, a doubt emerges as to whether one is exaggerating the importance of investor protections to the detriment of consumers, which can explain the excessively high rates charged by major, privatized corporations for publicly necessary services (e.g. telephony services in some of the countries under study). A similar risk involving competition is to be found in some sectors. in an effort to attract greater investment, policies on competition and concentrations are left lax. It would appear indispensable to understand that only strict policy can assure the desired benefits from foreign investment.

A very important feature of competition conditions in Central America is the influence exerted by international concentration processes on the level of production (telecommunications and banks, among others). This phenomenon has emerged out of the massive wave of mergers and takeovers that got under way in the 1990s and continues unabated. The technological advances that allows various sectors to achieve greater flexibility in production should help open the market to a greater number of competitors, but while that has been the case to a certain extent (wireless phone service and electric power distribution, to cite two examples among others), it has failed to adequately offset the aforementioned concentration process, and frequently it is the same major corporations that take advantage of such technologies to diversify production and in this way further strengthen their conglomerates. While this is by no means an exclusively SDC phenomenon, and is more of a generalized process, such countries require a special effort as they lack the sufficiently powerful competition-policy instruments needed to limit market power abuses by major corporations within their countries. Singh (2004), for example, suggests that the defence of competition in developing countries would best be conducted on a regional level in order to counteract the power of the aforementioned multinational firms, and argues that national authorities are too weak to apply competition principles inside each country in an independent manner. Singh proposes the need for a supranational body to regulate the behaviour of this type of firms and limit their ability to expand through mergers and acquisitions, the effects of which are predominantly anti-competitive. In this sense, he coincides to a certain extent with Adhikari and Knight (2004), who propose regional rather than national competition policies. Both positions deserve greater study.

In synthesis, the characteristics of SDC such as those of Central America that demand special attention in the formulation or reformulation of competition policies are as follows:

• The public's lack of competition culture and the heavy weight of the corporate organization of production on consumer mentality.

• Concentration of economic and political power in a few families, thereby complicating the independence of competition authorities and facilitating the capture of public entities by powerful economic interests. This is a characteristic closely associated with a precarious rule of law.

• Scarcity of financial and human resources for the creation of institutional structures strong enough to combat anti-competitive practices.

• An abundance of small and medium-sized firms and their lack of national and global competitiveness. They may achieve a more viable entry into the international market by forming alliances among themselves – export cartels – in which case competition law would have to be applied only to the anti-competitive practices of truly large-scale firms.

• Given the small scale of these markets, regulation is especially important for achieving a more proper behaviour on the part of the companies that operate within them. This is especially valid in the case of markets for non-tradable public utility services.

• The judicial system of the countries under study is extremely precarious and lacks the capacity necessary for efficiently managing cases that competition authorities might tackle.

• Each country's abilities to combat anti-competitive practices are weak in the face of the growing international trend towards a concentration of production (in telecommunications, banking and others), so regional cooperation is indispensable in this regard.

II. Economic Reform and the Rise of Competition Law in Central American Countries

Central American countries, as in the case of many others, promoted broad programmes of economic reform as far back as the 1980s, and with particular vigour beginning in the 1990s. The reform programmes in some of these nations lagged owing to serious problems arising out of acute political crises, which in various instances escalated into civil wars. Nevertheless, things began to regain a sense of normalcy beginning early in the 1990s and significant efforts to achieve economic reform got under way. During an initial phase, these policies came in response to the problems that the traditional model of development began to experience in recent decades, the onset of a new intellectual mood both locally and internationally and the pressures exerted by international financial bodies amid the foreign-debt crisis that erupted at the beginning of the 1980s. Such developments prompted countries in the region to try and find a new way to insert themselves into the international economy. Some of the reforms from this period were conducted in the absence of the sort of legal and institutional framework essential to their success. Those that were aimed at leaving resource allocation in the hands of the market and relieving the state of that burden did so without creating conditions that would have allowed these markets to function properly once the deregulations, privatizations and trade and financial openings occurred. Such weaknesses have heightened concern within all of the countries under study for the need to establish new measures for overcoming the limitations of the initial economic reforms. One such measure was the introduction of competition policies in the 1990s or in the middle of the current decade, which is to say less than ten years since the first wave of reforms were implemented.

As the current decade began, the countries of the region were at varying stages of progress on competition matters and offered a rich amalgam of situations (Table 2.1). In 2005, some of them (Costa Rica, Panama and Mexico) were already revising laws that had been put into place roughly ten years earlier, others (El Salvador and Honduras) had just adopted a law that drew on some of the accumulated experience of the first group of countries, and the last two countries, Guatemala and Nicaragua, were continuing to debate either a draft of a law or a final bill, respectively. Most of the laws and drafts of competition law built on Mexico's experience and legislation.

A series of elements, such as the insistence and support of international bodies, served as an impetus to competition policies in the region (see Schatan and Avalos, 2006). The heterogeneous response to the stimuli and/or demand from such international agencies depended on specific circumstances in each country including the power wielded by major corporations and the extent to which governments could function independently of power groups. Even in the countries that have competition laws and authorities, their characteristics (their strengths and weaknesses) depend on the degree of resistance exerted nationally by various sectors when the draft laws were being negotiated with the various groups. We can describe at least three situations in this regard within Central America:

• Countries that have implemented extremely broad economic reforms, including massive privatizations, and which have adopted competition laws and regulatory bodies. Panama has been in this situation since 1994, while El Salvador and Honduras joined this group in 2006. Honduras' privatizations have not been as extensive as in other countries, owing to difficulties encountered in passing control of some companies to the private sector, but the country remains firmly committed to that goal (Tábora, 2007). Panama's law was revised in 2006.

• Countries with privatization and trade liberalization, with weak regulations and lacking a competition law, such as Guatemala and Nicaragua. These countries are considering a draft of a law or a final competition law bill, respectively, with Nicaragua closer to adopting it than Guatemala as it is already before Congress (since July 2005) and lawmakers have already established some points of consensus on the proposal.

• Countries that have introduced economic reforms that are more limited on the level of privatizations, but which have adopted a competition law and agency. This is clearly the case of Costa Rica, which has a competition law, and where the

Table 2.1 Legal and institutional framework: competition and regulation (August 2006) (Based on official information)

state maintains a considerable presence on the level of banking, infrastructure, electric power and telecommunications, among other public utility enterprises. Its competition law, however, exempts public sector firms from its enforcement.

Costa Rica is the country that maintains the strongest public sector industry in Central America, one that includes a series of natural monopolies. Despite that, very early on the country embarked on an effort to reorient its growth model towards exports by introducing a programme for deregulating prices and the financial sector. Beginning in 1994, Costa Rica began to emphasize the search for free trade agreements.13 Approval of the country's competition law was motivated in part by interest in achieving the Free Trade Agreement with Mexico, which contained a clause on competition, and by the need to negotiate the Third Structural Adjustment Program (SAM III), whose conditions included having a competition law. When Costa Rica drew up its law, it patterned it after that of Mexico (Yong, 2005; Sittenfeld, 2006 and her chapter in this book). The extent to which the government enjoyed a greater degree of independence from the private sector compared to those of other countries in the region explains the relative ease with which the law was approved. However, the problem with Costa Rica's competition policy is that it contains numerous exceptions aimed at protecting public sector monopolies.

Different factors led to Panama adopting a competition law and establishing a Commission on Free Competition and Consumer Affairs (CLICAC). One of the principal motives underpinning this initiative was to achieve the institutionalism needed to support its joining the WTO in 1997. The Competition law (Law N° 29) enacted in 199614 contains the instruments for combating unfair trade practices

13 In 1994, after joining the General Agreement on Tariffs and Trade (GATT), tariffs were lowered to no greater than 55%, except on dairy products, poultry and their derivatives. Costa Rica adopted trade and reciprocal investment agreements with Mexico in 1994, Dominican Republic in 1998, Chile in 1999 and Canada in April 2001, the Caricom in March 2004 and is currently discussing a free trade agreement with the USA (DR-CAFTA). Costa Rica sustains reciprocal investment promotion and protection accords with more than 20 countries (Yong, 2005).

14 The Competition Law was enacted the same year as the Law on Industrial Property (Law 35), both of which are complementary.

internationally such as dumping as well as undertaking consumer protection. The law contains rules on four issues: monopolies, consumer protection, unfair trade practices and safeguard measures. The law emphasizes the effect of anti-competitive conduct on consumer well-being. In fact, the function of the CLICAC is frequently confused with that of setting prices favourable to consumers. The Panamanian government received considerable support from international financial bodies in preparing to deal with competition policies. This support came from the US Agency for International Development (USAID), which financed the initial studies for the draft of the competition law, and later the World Bank conducted a broad training effort. The new institutional arrangement that is expressed in a series of laws and the establishment of new bodies includes the 1996 establishment of the Public Services Regulatory Body (Ente Regulador), whose tasks include setting rates on public services that are in the process of being privatized (Fernández, 2005). Significant changes were made in 2006 to the competition law and the CLICAC was dissolved by decree to be replaced by the National Authority for Consumer Protection and Defense of Competition (ANPCDC). The changes made to the law included extending its application to all entities and types of economic activity with the sole exception of those that are of the unique purview of the state, and a major restructuring of the competition agency, which changed its character to a certain extent, as we shall see later in this text.

The case of El Salvador is different from the other countries we have mentioned. While proposed versions of a competition law were produced even before they first emerged in Costa Rica and Panama – such as a draft bill drawn up by the Salvadoran Foundation for Economic and Social Development (Fusades) in 1993–opposition from business groups and political parties in the Legislative Assembly assured that such bills would end up being filed away in the stacks of the Economic Commission. Interest in competition law revived in 1996, when the Center for the Defense of the Consumer and the Foundation for National Development (Funde) presented the Legislative Assembly with a second bill, and when in 1997, El Salvador took out a World Bank loan for implementing its National Competitiveness Program (PNC), which stipulated a strengthening of the institutional framework for the promotion of competition defence and consumer protection. USAID and the European Union were among the other bodies that worked with academic institutions in bolstering local capabilities in designing and analysing competition policies.15 However, not until evidence emerged that the economic model was stagnating did a possibility emerge for approving a competition law. In the wake of the major setback the governing

15 In 1998, USAID financed a project run by the Programa Académico de la Fundación Doctor Guillermo Manuel Ungo (Fundaungo) on strengthening competition conditions and the consumer protection system in El Salvador which identified good international practices in the design and application of competition policies as well as generating consensuses among key protagonists on the need for a competition law. In 2000 the European Union provided funding for the Friedrich Ebert Foundation to execute a programme for strengthening Salvadoran democracy that included training and technical assistance to the Legislative Assembly's Commission on Free Competition in an effort to assist in the process of designing and discussing a competition law.

Alianza Republicana Nacionalista (ARENA) Party suffered in the 2003 presidential, municipal and legislative elections, internal and external pressure began to build in favour of a recomposition of government economic strategy, and they ultimately achieved the necessary consensus on the content of the competition legislation,16 which was approved in late 2004 (Molina, 2007).

In Honduras, the task of introducing competition policy and, more specifically, a competition law has been made more difficult than in the other three mentioned countries owing, above all, to that country's significant degree of economic concentration and the intimate working relationship that exists between the government and the business sector, which rests on a web of deep personal and family ties that criss-cross sectors with economic and political power and ultimately influence many political and economic decisions (Tábora, 2007). Nevertheless, the country needed to make progress on a number of fronts before it would be capable of applying a competition law, especially in administering justice. For example, Honduras lacked a public prosecutor until 1993 when the Law on Public Prosecutors was enacted. It was impossible to effectively implement the Consumer Protection Law approved in 1989 until the 1997 establishment of the Special Consumers' Prosecutor's Office (attached to the Public Prosecutor) for prosecuting cases. Another series of laws necessary for applying competition law was recently enacted in Honduras, including an investment promotion law, the Law for the Promotion and Development of Infrastructure and Public Works (1998), a revised version of the Law on Industrial Property (2000), among others (Hernández and Schatan, 2006). Each of these measures helped to establish the principles under which agents could exercise their economic freedom in the broadest possible sense. But major resistance to the competition law remained, and the consensus necessary for approval failed to emerge despite the first efforts to draw up a version of the law, dating back to 1994 when the Office of the Presidency prepared a discussion document (Guidelines for formulating and implementing a law for promoting economic competition), and in 1996 when the first draft was published.

Renewed support for the idea emerged when the idea was listed as one of the commitments in a Poverty Reduction Strategy accord signed with the International Monetary Fund (IMF), and again among the 2004 political goals assumed under the Poverty Reduction Strategy for Honduras. In 2002, the National Competitiveness Commission was formed, whose functions included furthering the development of the PNC. The latter's tasks included improving competition and lowering administrative barriers, thereby giving added impetus to approval of a Law for Promoting the Defense of Competition. These endeavours enjoyed special funding from the World Bank. The competition law was finally enacted in February 2006 and the agency was established in August 2006.

16 The main points of consensus that emerged included the need for an independent competition agency and on sanctions that would discourage anti-competitive conduct, but which would not be confiscatory for firms.

In Nicaragua, the government remained wedded to promoting a strong public sector economy until 1989. When we add to this hurdle the economic crisis and the worsening social conditions with which the country entered the 1990s, we can appreciate just how much more of an effort would be needed to generate conditions favourable for achieving conditions of effective competition than in other countries in the region. A number of key reforms were enacted once the country emerged from a period of hyperinflation by 1991 and growth resumed in 1994. Cooperation programmes were resumed in 1992 with a series of donors – the United Nations Development Program (UNDP), German Technical Cooperation (GTZ, for its German-language initial abbreviation), the UK's Department for International Development (DFID), among others – when a support and advisory programme was created to assist with the competitive transformation of the country. In addition to macroeconomic issues, the programme anticipated a modernization of the legal framework governing the private sector as well as the institutional support upon which it could draw. Significantly, in Nicaragua it has been the forums for negotiating international trade agreements that have pushed the country to commit to, and enforce, competition laws and norms. Added urgency to approve a competition law has arisen from membership in Dominican Republic – Central America Free Trade Agreement (DR-CAFTA), which contains no explicit clauses in this regard but indirectly takes up competition in the case of the banking and telecommunications industries (Ansorena, 2007). The Nicaraguan Congress was debating the law in mid-2006 and the small and medium-size enterprises were actively supporting the law project.

Guatemala has expressed the least interest in enacting a competition law of any country in the region. The need for constitutional and legal norms in defence of competition has not been a topic of significant debate in forums within civil society or business circles; nor has it become an important item in the agendas of either political currents or the government (Romero and González, 2006). Since the second half of 2001 a draft version of a law has languished in the Economy Ministry's Department for the Promotion of Competition, but it took years to emerge as a bill. That lack of progress apparently reflects the extent to which the individuals who alternated as ministers of the economy from 2000 to 2004 were unconvinced that the proposal might win the necessary support either in the executive branch or in Congress (Romero and González, 2006). Furthermore, some leading government officials were also businessmen with direct interest in some of the most protected segments of the economy, another factor that further inhibited officials at the Economy Ministry from more enthusiastically promoting the draft law. One major source of resistance to even discussing competition law in Guatemala, as in the rest of the countries under study, is the extent to which monopolies or oligopolies control domestic markets for a number of strategic goods and services.

In synthesis, a long list of factors have led to a protracted process in approving competition law (El Salvador and Honduras) or kept such legislation from yet being approved by mid-2006 (Guatemala and Nicaragua). Those obstacles include resistance on the part of business interests that have long been protected or which have enjoyed exclusive concessions or acquired state-owned companies at very favourable terms; the influence such sectors wield within government; the very marginal extent to which this matter has been incorporated into the economic agendas of governments; a lack of local technical capacity to seriously and thoroughly explore the issue; and a paucity of competition culture in these countries. Even those nations that enacted competition law and established competition authorities in the mid-1990s and which have managed to implement reforms that restrict market power abuses continue to face hurdles. They continue to run up against the resistance posed by the most reticent power groups to effective competition policy; they feel pressured to draft their laws with too great a haste or pattern their legislation after those of other countries while failing to account for the peculiarities of Central American nations; or they lack the human capital and necessary funding to carry through on this task, or any combination of the above factors.

III. Competition Norms and Institutions in the Laws and Draft Bills Under Discussion in Central America and Mexico: A Critical Evaluation

1. Characteristics of the Competition Laws in Central America and Mexico

Concern regarding the eventual negative effects from monopolies can be found in all national constitutions in the region and, while they vary in form, all of them prohibit monopoly and anti-competitive practices. However, for an extended period of time they lacked corresponding regulatory laws, meaning that such constitutional principles were dead letter in practice. However, the existence of such nominal rules aided governments or other interested parties in promoting a legal and institutional competition framework. As we have already indicated, most Central American countries already have competition laws and regulatory bodies, and those that do not have drawn up competition bills and created offices in charge of promoting their adoption and implementation. After Honduras approved its law in early 2006, the formation of a Competition Authority was promoted by the PNC, which in turn formed part of the Foundation of Export Research and Development (FIDE, for its Spanish-language initial abbreviation). In Nicaragua, the Ministry of Promotion of Industry and Commerce (MIFIC) has a General Department of Competition that directly oversees market regulatory issues, especially those related to competition. Guatemala's Economy Ministry has a Deputy Ministry for Investment and Competition (which includes a department on competition), which is drawing up the law and norms designed to favour free competition and the elimination of protectionist practices (Romero and González, 2006). Nevertheless, the agencies in Guatemala and Nicaragua that have been assigned the task of promoting competition law lack the decision-making power needed to define market rights. As a result, they are limited to generating information and lack the ability to embark on an active policy of competition supervision.

In the following analysis, we compare the existing laws and bills under discussion. We seek to identify the various national perspectives and both the similarities and differences between countries on the level of competition. In addition to the six countries of Central America, we take up the case of Mexico, whose laws inspired those of several countries in the region.

An initial, general overview appears in Table 2.2, which lists the dates on which competition laws were approved, the name of the law, the name and date on which the respective government agency began operations, and whether the law is or is not under review.

Most of the existing laws or proposed competition laws in Central America define their main purpose as promoting competition as a means to improve economic efficiency, and in this way to raise consumer well-being. Something different happened in Guatemala and Mexico; in the former country the argument is that efficiency will benefit both producers and consumers, while Mexican laws make no explicit reference to benefits for either producers or consumers. Countries in the region tended to adopt the European model of establishing norms for the anticompetitive conduct of economic agents more than impeding the emergence of monopolies or oligopolies in domestic markets. This perspective calls for a legal framework along with a body of norms that are universally applicable to all who participate in the market for goods and services.

The classification of anti-competitive practices in competition laws differs from country to country, but the concepts are quite similar. Following the example of Mexico, Panama and Costa Rica distinguish between absolute monopolistic practices, relative monopoly practices, concentrations and unfair competition. In El Salvador, the law lists competitors, agreements between competitors, agreements between non-competitors and concentrations. However, the concept of agreements between competitors is fully applicable to absolute monopolistic practices, while agreements between non-competitors are typified as relative monopolistic practices. Guatemala's typification identifies prohibited behaviour, abuse of a dominant position and concentrations. An interesting concept in the Guatemalan draft law is that of functional competition, which refers to the difference between real competition, conditionally restricted practices and concentrations. The Honduran law distinguishes between practices restricted by their nature and those that are prohibited due to their effects. Although the terminology differs, this is another instance in which the law is speaking of absolute and relative monopolistic practices. In Nicaragua, the typification is that of monopolistic practices (concerted practices and anti-competitive agreements) and abuse of a dominant position as well as unfair competition. Concerted practices and anti-competitive practices refer to absolute monopolistic practices while abuse of a dominant position refers to relative ones.

Three countries under study – El Salvador, Honduras and Mexico – have defined specific antitrust thresholds beyond which competition is harmed and at which point official notification is required before any mergers or takeovers are entered into. The other countries lack any such notification rules but companies may voluntarily give advanced notice and officials are free to retroactively review such

Table 2.2 Laws, draft bills and competition commissions in Central America and Mexico 1992–2006 (Authors, based on official data)

concentrations and order a break-up of the firms in question ex post when they deem such measures are needed (see Table 2.3).

There are significant differences between the laws in terms of the areas that each addresses and of the responsibilities and functions assigned to competition authorities. The Panamanian law and the Nicaraguan draft, for example, list numerous tasks in addition to the defence of competition. Consumer protection, unfair practices in international trade and, therefore, antidumping and import safeguards are also included in the Panamanian law.17 Under the Nicaraguan bill, the competition authority must deal with market competition, consumer protection, intellectual property, simplifying filing and application processes and unfair competition (in the sense of misleading or deceiving consumers). In the other countries competition laws and draft versions exclusively focus on competition, except in Costa Rica, where consumer protection is also mentioned, but two separate entities were created for each subject: the Competition Promotion Commission and the National Consumer Commission.

Both existing laws and pending legislation generally extend competition law to all economic agents whether public or private. However, exceptions abound. In Mexico, Costa Rica and Panama, the laws do not apply to monopolies that were legally created for industries that are constitutionally reserved for the state. Many of the exceptions that were once contained in the Panamanian law were eliminated during the 2006 revision, leaving only providers of public utilities (electricity, water, etc.) under the scrutiny of the Competition Authority. Guatemala's draft law and Nicaragua's bill, as well as existing law in Costa Rica and Mexico, allow for export cartels with some conditions: that they neither reduce the value of exports nor exclude other agents involved in the same type of economic activity. Costa Rica's competition law also does not apply to service providers whose activities are regulated by concession contracts. The pending Guatemalan and Nicaraguan laws would exempt their application to intellectual property rights. In Mexico, Honduras, Nicaragua and Panama, competition law does not apply to collective bargaining agreements. The Costa Rican law contains the greatest number of exceptions, but the recently designed laws or draft versions of such laws in other countries tend to be much more broadly applicable. Practically no exceptions are to be found in the laws of El Salvador and Honduras, for example, and in the latter of these two countries the law extends even to economic sectors regulated by special sector laws. The Honduran law is the only one in the region that even applies to foreign-based professional and business associations whose actions may impact the local market; not even Mexican law contains such clauses.

Some limits on spontaneous market functioning arise out of the government's authority to set some prices. Mexico's law leaves open the possibility of setting prices of products that are regarded as essential or basic consumer items. In Costa

17 The 2006 revised version of Panama's competition law also created separate offices for consumer and competition protection.

Table 2.3 Thresholds for analysing concentrations (The competition laws and draft bills of each country)

Rica the law allows for regulation of rates on public utilities that are still provided by state-owned companies, as well as to ration import and export licences for some products on behalf of public interest, but this mechanism can be employed only on a temporary basis and the arguments cited in its application are subject to review every six months. The Guatemalan draft law states that it does not take precedence over other laws or their accompanying rules and regulations (Article 5) that are aimed at objectives such as adjusting supply, promoting exports and dealing with idle capacity or social issues.

One of the most difficult tasks facing competition authorities is that of market definition for antitrust purposes. Five out of the seven countries in our study use the same criteria for defining a "relevant market" (Panama uses the term "pertinent market"). The relevant market normally refers to the product market and the geographical market. The former refers to the market for goods that respond to a certain specific demand that can be satisfied with similar, interchangeable and comparably priced products, while the second market refers to the area in which these goods may be obtained on favourable terms (e.g. at a distance that would not imply an additional charge for transportation costs). When it comes to sanctions, the statutory maximum fine in Mexico for absolute monopolistic practices can be as high as $1.55 million and the one for relative monopolistic practices as great as $935,000.18 In 2006, the highest potential fines in Central America were in Panama, at $1 million, followed by El Salvador, where they can reach $800,000, depending on the seriousness of the infraction. The other countries contemplate fines that are much smaller in absolute terms. Fines defined as a percentage of earnings or capital might prove severe, but these practically have never been imposed. This type of fine is contained in the laws of Costa Rica and Mexico. In this last country, such sanctions can be particularly severe as they can be as great as 10% of the annual revenues the guilty company took in during the preceding fiscal year or up to 10% of its asset value, whichever is higher. The draft law in Guatemala would impose fines of up to 100% of sales from the preceding year in extreme cases. The recently approved competition law in Honduras also leaves open the possibility that companies can appeal such rulings by filing a writ of amparo.

There are many obstacles to imposing the penalties contemplated in competition laws. In Mexico, companies accused of monopolistic practices or abuse of market power can easily appeal such rulings, thereby postponing the levying of fines or other penalties for a long period of time (Avalos, 2006). In a similar vein, there are a number of means in Panama for avoiding fines (only 10% are paid) or other punishment, at least under the law as it read prior to the revision of 2006.

As for non-monetary penalties, all of the laws or proposed laws in this study contemplate the possibility that competition authorities may order that the prohibited activity be ceased or corrected, including the partial or total break up of a business that has been improperly merged. Except in Nicaragua, all such laws allow for the possibility of filing criminal charges whenever the authority has exhausted all administrative recourse or such action is warranted by the seriousness of the accused party's conduct. Nicaragua would only impose fines and those are to be calculated based on what is deemed necessary to compensate for, or repay, the damages suffered by the affected parties. While El Salvador imposes sanctions

18 These would be greatly increased under the revised version of the competition law that the Mexican Congress was considering in mid-2006.

for procedures that contravene the law or regulations, the penalty may be lifted if the company cooperates in correcting the problem under a procedure we can generically refer to as a leniency resource. This option provides authorities with means for obtaining essential information on anti-competitive practices that they might otherwise have a hard or impossible time obtaining. The proposal to reform Mexico's competition law also introduces this concept, which is already widely used in many other countries including the USA, where companies that self-report violations or offer early cooperation with criminal or regulatory investigations may avoid criminal prosecution and fines.

Competition advocacy is a crucial issue. It is conducted in various ways in a number of countries. In Mexico, advocacy extends both to the promotion of a competition culture and the issuing of opinions – either on request from the federal government or by the authority's own initiative – regarding new laws or norms, and if the revised version of the law is approved such opinions would be binding on such rules. Costa Rica also allows the competition commission to consider whether application processes, requirements and regulations on economic activity obstruct, impede or distort either domestic or international market transactions. The Nicaraguan draft law contains a section on competition culture and refers to the simplification of filing procedures, reviews of judicial rulings and the treatment of government assistance. This is the only group of countries covered in this study that textually confers on the competition commission the power to specifically engage in competition advocacy. The laws of El Salvador and Honduras and the draft competition law in Guatemala make no mention of this subject.

2. Competition Agencies

Although there are substantial differences between the competition authorities of the countries covered by this study, something of an international consensus has formed as to the characteristics a competition agency needs in order to function efficiently and effectively. The most significant of these features include the need for autonomy from the executive branch of government; stable funding; that commissioners meet certain ideal standards (that they be competition specialists, that their nominations be widely accepted, that they serve a clearly defined term of office and that they be subject to removal for reasons that are clearly spelt out in the law). It is also recommended that the process of choosing and replacing commissioners be staggered so as to better assure the possibility of consistent rulings and an independent outlook.

As Table 2.4 shows, under the laws of Costa Rica and Mexico, the competition authority is attached to the Economy Ministry, which draws up its budget. The competition agencies of El Salvador, Honduras and Panama send either to Congress or to the executive branch their own budget proposals, which are delivered by a ministry, usually the Finance Ministry.

Table 2.4 Institutional characteristics of competition authorities in Central America and Mexico (Competition laws and law projects of Central American countries)

In the countries under study most laws or bills initially opted for a commission rather than a single commissioner except El Salvador, where the authority is a superintendency presided over by a single director. In Panama's 2006 reform of Law 29, the three-member CLICAC commission was replaced by the National Authority for Consumer Protection and the Defense of Competition presided over by a single administrator.19

The commission formula offers the advantage that decisions are not made by a single individual but instead are the product of collective deliberations among various commissioners. A multi-member body may also prove less vulnerable to control by private economic agents. Nonetheless, the commission may be at a disadvantage in some respects compared to a single director, who might handle cases more expeditiously, for example.

The manner in which commissioners are chosen is another factor that determines the commission's degree of autonomy. A selection process, in which both the Presidency and Congress are involved, for example, may bolster a competition agency's autonomy and legitimacy by limiting the chances that commissioners could be arbitrarily removed. Panama's Law on Competition, Guatemala's draft law and Nicaragua's bill all contemplate such a selection process. In any event, commission autonomy must be accompanied by a proper mechanism for rendering accounts.

Due to the predominance of political systems that concentrate power in the Presidency, some countries have preferred to leave the faculty for designating competition authorities strictly with the president: Mexico, El Salvador and also Costa Rica, which added the step of having the Ministry of the Economy, propose a group of candidates from which the president chooses. In other countries, proposals by the president or a commission to designate candidates must be ratified by the Congress or Legislative Assembly (Guatemala, Honduras, Nicaragua and Panama). In almost all countries the designation of competition authorities is staggered (when there are appointments) or does not coincide with presidential periods, in order to isolate them from political pressures. Finally, in some countries these authorities may be re-elected and in others they may not (see Table 2.4).

Some legislation establishes norms aimed at ensuring that the individuals appointed to competition commissions have specialized knowledge (see Table 2.4). Although the small size of most of the countries considered in this study makes it difficult, at present, to demand professional preparation and experience in the field of economic competition, the laws require that the commissions include people with a background in economics together with professional lawyers. This is true of the recently approved Salvadoran law and the proposal under study in Guatemala. The case of Mexico is, without a doubt, the most advanced in this field, requiring

19 The new agency has two departments: the National Department on Free Competition and the National Consumer Protection Department. The administrator maintains a close working relationship with the five-member Council of Advisors, which is made up of the Minister of Trade and Industry, who presides over the council; the Economy and Finance Minister; the Minister of Health; a representative of the consultative council of the consumer associations; and a representative from the associations of business owners, retailers and/or industrialists.

that the professionals participating in the commission have experience in the area covered by the law.

In all the countries studied it has been made very difficult to remove competition authorities from office in order to prevent political interest groups from influencing the competition body. In Mexico, members of the commission can be removed only for well-founded causes, and other countries go one step further by specifying the only acceptable causes for removal (see Table 2.4).

Despite the multiple mechanisms that contribute to greater independence of competition authorities, budget limitations, political regimes and other factors in developing countries make it difficult to achieve total independence for the commissions and their members.

Last, it is important to note that competition legislation has been changing, or is in the process thereof, in the countries that pioneered competition policy in the region (Costa Rica, Mexico and Panama). These revisions respond largely, but not exclusively, to the difficulties encountered during the first decade of existence of the legislation.

Panama's Law on Competition was modified in February 2006. This new vision expands the law's reach to include companies that provide publicly useful services, as well as substantially altering the commission's internal structure in order to expedite case resolution, and raising fines and facilitating their collection. The new law also offers greater protection for competition authorities, who previously could face prosecution as a result of litigation from existing anti-competition cases. The new law, however, has become a source of controversy as the agency administrator has lost some of the autonomy previously enjoyed by CLICAC commissioners.

Costa Rica, for its part, was in the process of revising its competition law in 2006, but by mid-year a proposal for an alternative law was yet to emerge. The changes being debated were: improved typification of anti-competitive practices; possible requirement of prior notification of concentrations, with adequate establishment of limits; the incorporation of "clemency appeals;" the elimination of various exceptions to the application of the law; and the strengthening of institutions linked to competition policy, among others.

In Mexico, the latest version of the law (approved by Congress in mid-2006, but yet to be signed by the president as of this writing) grants greater powers to the Federal Competition Commission (FCC) to break up companies, including those in the telecommunications, media and financial services sectors, that engage in anticompetitive practices and fail to abide by earlier commission rulings. The new law imposes considerably higher fines than those imposed under previous rules in an effort to more seriously dissuade potential violators of the competition law. For example, the fine for instances of absolute monopoly practices would reach as high as 1.5 million times the daily minimum wage in effect at the time in Mexico City. Additionally, the "appeal for pardon" would be introduced to help the FCC bring together evidence to identify any anti-competitive practices by companies. Collection of the fines, which had been minimal (approximately 14% of the total levied), would be carried out by the Tax Administration System, which would be likely to increase the effectiveness of collections.

Within Central America, Costa Rica and Panama are the countries that have the most experience in resolving cases filed before competition authorities, given that in the other countries the competition commissions are very new (El Salvador) or do not yet exist. In Costa Rica, between 1995 and 2004, the competition authority resolved 537 cases, or 89% of those brought before the commission. Among the cases resolved during this period, there were 145 general consultations, 203 suits, 92 formal investigations, 47 consultations to the Technical Unit, 9 rulings on licences, 37 opinions, 2 instances of price setting and 2 mergers. Fines have tended to increase from around $2,400 in 1998 and to surpassing $400,000 in 2002. Nonetheless, the overall amount of fines applied is still quite limited. Through mid-2006, no penalty had yet been issued and paid that was equivalent to 10% of the value of annual sales or 10% of the value of a firm's assets (Yong, 2005; Sittenfeld, 2006, and her chapter in this book).

In Panama, CLICAC (whose name was changed at the outset of 2006) handled few cases of anti-competitive practices. In fact, between 1998 and 2004, approximately 30 cases of monopolistic practices were fully processed including both those brought by private companies or CLICAC.20 As was discussed earlier, the competition authority only has the faculty to investigate the existence of indications of anti-competitive practices, and if they appear to exist, the authority must file suit in the courts and wait for them to resolve the case. Since the CLICAC was created, 14 cases were settled (through 2005) without the need to file suit in the courts, either because it was decided to handle the case through competition advocacy or because evidence of wrongdoing was not sufficient to file suit. The rest of the cases were taken to court and by mid-2006 decisions had been issued at the initial and appeals levels; these cases involved meat, flour and medical oxygen, among other products. It should be noted that once a case is taken before the courts, and the judge rules in favour of the competition authority, the accused may appeal before the Superior Court and other courts, such that a case may be prolonged significantly (this has changed somewhat since the Competition Law was changed in 2006). It is interesting to note that in cases of absolute monopolistic practices brought by CLICAC that have been resolved, the courts have generally supported the position of CLICAC.

The majority of the 1,450 cases processed between 1993 and 2004 by Mexico's FCC (Avalos, 2006) were related to mergers or acquisitions, which can be explained by the requirement that the companies provide advance notification of these operations. This suggests that from the time this prior notification was required, the competition commission has had to devote considerable resources to resolving these cases. It should also be noted that the majority of the cases of monopolistic practices processed in Mexico have been of the "vertical" type (Avalos, 2006), in contrast to those handled in Costa Rica which have been principally "horizontal," partly because it has been harder to prove "vertical" practices.

20 Some cases are investigated confidentially and the number of such cases is unknown.

IV. The Competition Law and the Judicial System

International experience shows that in various developing countries, two principal mechanisms have served to debilitate the institutionalism of competition defence: the creation of weak competition bodies, and the establishment of weak or nonfunctioning judicial mechanisms that instead of assisting the commissions in applying competition law effectively neutralizes and weakens the law even more.

Application of competition policy obligates the state to intervene in transactions between private parties (or between private parties and the public sector) to preserve the competitive conditions of markets. Although diverse institutional models exist, the application of competition law, including the application of monetary sanctions, generally has been, or is to be undertaken by, an administrative body, either within the executive branch or in the form of an autonomous agency. In Central America, the exception is Panama, where the competition authority (ANDCC) may not apply sanctions until there is a court ruling. Other countries resolve cases through administrative channels, or they will once the law is passed. However, it will be possible to continue in this mode only if the parties found in violation of the law comply with their penalties. If not, the competition authority will resort to judicial channels once administrative options have been exhausted. Furthermore, if the sanctioned company feels that the penalty it has received is unfair, it may appeal or seek an injunction before the judicial system in Mexico and Honduras, thus overlapping the administrative and judicial systems. In some countries, once the competition authority makes its ruling, it may be appealed exclusively within the administrative system. At that point there are no more administrative channels and any successive litigation is handled by the court for administrative disputes. When the competition authority conducts its investigation, it may solicit information from the companies involved, and the majority of competition laws, or proposals for laws, grant the commission the power to demand this information; some even give the competition authority the power to carry out inspections in situ. Information provided by the companies is expected to be trustworthy, and if a company is found to falsify or omit any requested information, it may be subject to fines according to the laws of some countries. If these fines are not paid, competition authorities may seek redress through the judicial system. El Salvador imposes the lightest fines for this type of infraction, but it does impose them. In Mexico the provision of information is voluntary, but as the law is revised this will probably change, and in Panama, as discussed, the judicial branch intervenes from the start, issuing the instructions to provide information.

To the degree that competition authorities depend on judicial tribunals to carry out their functions, they face various obstacles to the effective resolution of their cases. Noteworthy among these are the following: (i) generally, judges are not prepared to handle cases of great technical complexity; (ii) the existence of administrative bodies that can issue legal rulings creates different options for the resolution of cases. In international practice at least three methods may be identified. (i) Perhaps the most common one holds that the administrative authority is just one party on an equal footing with the private party in the case, and in which the object of the dispute receives no special treatment. In consequence, cases are presented by the administrative body and are subject to all ruling bodies to which other legal cases would be subject. (ii) A second method recognizes the administrative authority's faculty to issue rulings, and in its capacity as a representative of the public interest, on certain differences with respect to the private party involved in the case. In this context, the role of the judicial system's review is reduced to a review of the due process carried out by the high court. (iii) A third method distinguishes between the administrative faculties of government bodies and autonomous agencies and the courts' faculties for issuing rulings, but recognizes the special characteristics of competition cases, and in consequence, creates special tribunals for them. The basis for this method is that the judges can become specialized in the issues relevant to competition cases, allowing them to manage the technical complexities involved. At the same time, this guarantees greater coherency and consistency in decisions and presumably allows for swifter resolution of cases. The countries in the study group with competition legislation and authorities use various methods. In Costa Rica, El Salvador and Honduras, cases brought before the competition authority are resolved by that same body, which has the faculty to apply penalties. As previously noted, parties have recourse to an appeal or reversal of the authority's final resolution. As such, once administrative channels are exhausted, the legality of the final resolutions may be disputed directly before the courts of administrative disputes.

In Panama, the National Authority for Consumer Protection and the Promotion of Competition (previously CLICAC) has the characteristic of being an investigative body that does not issue binding rulings. This agency only reports on evidence of any anti-competitive practices; if an investigation leads it to conclude that there are grounds for proving anti-competitive practices, it has the option of presenting its case before the courts. For this purpose, Panama created the Civil Circuit Courts,21 which are responsible for resolving cases resulting from the application or interpretation of the law regarding monopolies, consumer protection and bad-faith business practices. These courts were intended to overcome the difficulties faced by the ordinary courts in resolving highly technical cases and, at the same time, to ensure swift resolution. Few cases have been brought before these courts, however, and resolution of these cases has been extremely slow.22

21 In the 1970s, once new laws for business were defined, Panama debated whether these laws should be interpreted and ruled upon within the administrative system, in the traditional courts or in specialized courts. International organizations and the executive branch preferred that these issues be resolved administratively, as was taking place in the commercial arbitration tribunals of the Chamber of Commerce. In 1978, however, a decision by the Supreme Court regarding commercial exclusivity contracts determined that commercial issues should be taken up by the ordinary courts and should not be resolved through administrative channels. Following passage of the Industrial Property Law and the Competition Law, the volume of business justified the creation of the Specialized Commercial Courts (Fernández, 2005).

22 In part, this situation can be explained by the fact that during the period 1998–2002, these courts received 2,511 cases, of which 93% were suits regarding intellectual property rights (Fernández, 2005).

The Third Superior Court of Justice of the First Judicial District was also created and made specifically responsible for handling appeals of rulings handed down by the circuit courts.23 In practice, the authority brings a suit before the court either based on a complaint by third parties or ex officio in a summary oral hearing. The case is brought before the respective Commercial Court, which sets a preliminary hearing. Subsequently, an in-depth hearing is set for the oral presentation of evidence by any of the parties. Once the judge issues a decision, the decision may be appealed to the Third Superior Court. In the case of rulings above 750,000 balboas, the affected parties may seek a higher appeal before the First Hall of the Supreme Court of Justice, but when penalties are below this amount, no appeal is allowed (Fernández, 2005).

In Mexico, the Federal Competition Law allows for an appeal for reconsideration before the competition commission (Article 39). This appeal suspends execution of the resolution while the commission considers the appeal request (within 60 days). This first phase extends the period of action taken by the commission, leading to a second stage that begins once fines are issued. The competition law establishes fines that appear to be sufficiently significant to dissuade or punish parties engaged in anti-competitive practices, but the fines ultimately are practically unenforceable due to the many defence mechanisms available to companies. The principal problem results from the use of the writ of amparo, originally designed to protect individual guarantees as a means of resolving disputes within administrative law. Through April 2004, which marked the tenth year of the Federal Competition Commission, 6,666 cases were resolved, against which 636 amparos and 90 rulings of fiscal nullification were applied. Among these, as of the date indicated, 375 amparos had been resolved, with 260 still pending. Of these injunctions that were resolved, 35% was decided within one year; 76.5% was in less than two years and the remaining 23.5% in a period of three to five years. Of all the amparo filings, 284 were sought against procedural movements to protect companies from requests for information or allegations of responsibility for anti-competitive practices (Sánchez Ugarte, 2004).

The amparo presents various problems when applied to companies. It does not seem reasonable for a legal recourse aimed at protecting the life and safety of individuals to be used in commercial suits. Even more so when the accused in question are alleged by the Commission to be engaged in prohibited practices that gravely affect smaller companies, as is very frequently the case. Furthermore, the suit of amparo does not explicitly take into account the protection of the public interest. A second problem arises because of the very nature of the amparo, because the case becomes focused on the protection of guarantees with little attention paid to resolving the original dispute. We can cite a series of additional complications: the possibility that each and every subsidiary of the company obtaining the amparo also seek protection,

23 Panama also has municipal courts that handle only cases involving claims of 3,000 balboas or less on the part of the consumer. In the provinces or districts that do not have one of these courts, cases involving free competition and consumer issues are handled by the civil or mixed circuit and municipal courts.

and given that such processes are not cumulative, each of the subsidiaries' cases may be brought before different judges and courts and result in different and even contradictory rulings. Such complications are further magnified by the absence of any time limits for courts to issue decisions on such cases (Sánchez Ugarte, 2004).

How can these problems be addressed given that they severely limit the possibilities for effectively defending competition? In Mexico various measures have been proposed that tend to restrict the use of the amparo. These include combining all the cases brought over the same issue; the introduction of greater requirements for the granting of such an injunction24; that ruling on amparo requests take into account all the presumed misdeeds and issues of both form and substance25; requiring the courts to rule on such cases within a specific period of time; and assuring that the sentence be applicable to all economic entities. Several of these aspects have been incorporated into the proposal for revision of the competition law approved by Congress (mid-2006) but pending approval by the president.

During the discussion surrounding the elaboration and approval of the competition law in Honduras, one bill proposed that once administrative channels are exhausted, a case may proceed to the courts on the condition that the accused must already have paid the fines levied. The logic behind this proposal was to counteract actions meant to prolong the process, thereby undermining the effectiveness of the law. Nonetheless, this element was removed from the latest version of the proposal. The bill now stipulates in its Article 48 that once administrative appeals against the decisions of the commission have been exhausted, the only further recourse will be to seek an amparo before the Supreme Court of Justice, which could generate problems of the type mentioned in the case of Mexico.

The case of Chile is illustrative. Based on experience accumulated by the traditional system,26 Chile created the Court of Defense of Free Competition as a special and independent ruling body, the function of which is to prevent, correct and sanction actions deemed prejudicial to free competition. The Supreme Court is in charge of the directive, correctional and economic superintendency of this court. In this way, the court assures the existence of personnel who have been trained to

24 On this point it has been proposed that judges not grant an amparo when the ruling in question is against a monopolistic practice that substantially affects the public interest, and that in cases where an amparo might be issued, such a ruling could only come after the party seeking the injunction assures future payment of damages to the harmed party through a bond or other adequate guarantee (Sánchez Ugarte, 2004).

25 If this proposal was to be approved, as reasonable as it seems, it would definitively transform the entire concept of an amparo.

26 Until the reform of November 2003, the Chilean anti-monopoly system consisted of the National Competitiveness Enforcement Department (Fiscalía Nacional Económica), an administrative body and both prevention and resolution commissions. The system recognized the need for specialized personnel (in this regard, it was not constituted by judges but rather included economics professionals and government officials) but even as the Resolution Commission was presided over by a Supreme Court judge, its rulings were only subject to review by the country's high court, ensuring agile dispatch of cases. The Court of Defense of Free Competition recently replaced the two commissions mentioned.

tackle the difficult technical problems of competition legislation, consistency in rulings and, at the same time, swift resolution of cases, given that decisions may be appealed only before the Supreme Court. A formal complaint may be filed in the case of definitive rulings on measures included in Article 17 K,27 and such a request can be made either by the National Competitiveness Enforcement Department (Fiscalía Nacional Económica) or by any of the parties involved within a period of ten working days from the time of notification. In order to ensure swift processing, the law establishes that this claim be given preference over other issues. Furthermore, to assure termination of the anti-competitive conduct, the law establishes that the filing of a formal complaint will not suspend compliance with the decision except for the payment of fines.28 However, at the specific request of one of the parties and via a well-founded resolution, the court in which the claim is filed may totally or partially suspend the effects of the sentence.

V. Market Distortions and Cases Resolved by Competition Agencies

The objective of this section is to use examples to illustrate market distortions that occur throughout Central America and Mexico. National studies carried out as part of the project "Reinforcing Competition in Central America" by the International Development Research Centre (IDRC) and Economic Commission for Latin America and the Caribbean (ECLAC) reveal that in certain markets, similar anti-competitive practices repeatedly occur. Some of these practices are part of regional or even international business strategies, and others are exclusively domestic in scope.

Setting aside the telecommunications and banking sectors, which are developed in other chapters of this volume, sectors such as cement, rice, milling, beer and other alcoholic beverages, air transport, the poultry industry, slaughterhouses and beef are particularly anti-competitive, although some of these have recently undergone a positive evolution. In the following section we will analyse three markets: the milling industry, in which market distortions occur at the national level; the abuse of market

27 The measures included in the Article are the following: (i) to modify or put an end to the acts, contracts, pacts, systems or agreements that are contrary to the dispositions of the present law; (ii) to order the modification or dissolution of associations, corporations and other private corporate entities that may have intervened in the acts, contracts, pacts, systems or agreements referred to under the previous letter; (iii) to apply fines payable to the public fund of a sum equivalent to up to 20,000 annual tax units. The fines may be imposed upon the corresponding corporate entity, or its directors, administrators and anyone who may have participated in the conduct of the respective act. In the case of fines applied to corporate entities, their directors, administrators and anyone else who has benefited from the respective acts must be responsible for payment, as long as they have participated in the conduct of these acts' (own translation; Republic of Chile, 2005).

28 To file a formal complaint in cases in which a fine is imposed, the guilty party is required to post a sum equivalent to 10% of the fine.

power by the airlines at the regional level; and anti-competitive practices in the soft drink industry in which global companies are involved.

1. The Wheat Flour Market

The flour market is of great concern to governments for its wide-ranging social impact on the economy (representing 7% of the basic food basket in Panama; Fernández, 2005).

This sector has been analysed in the cases of El Salvador (Molina, 2007), Nicaragua (Ansorena, 2007) and Panama (Fernández, 2005), but the problem also exists in countries such as Costa Rica and probably in the rest of the region. In the first half of the decade of the 2000s, the problem of collusion among flour millers became more evident and severe as the price of wheat rose on international markets, affecting the price of flour and subsequently of bread at the national level. In Nicaragua, for example, between 2001 and 2004 the price of flour rose 55% (in fact, there was a bread "crisis" in 2001 in this country).

Generally, the flour market has a monopolistic or oligopolistic structure in which the key players also often retain control over flour imports and distribution (creating distortions such as market segmentation). Some countries even depend on imported wheat for domestic flour milling.

In El Salvador the milling industry is duopolistic: Molinos de El Salvador (MOLSA) and Harinas de El Salvador (HARISA) control the three phases discussed. In Nicaragua there are only three flour millers, and they maintain very close ties to the two flour importers Proharina and Fhacasa.

In Panama, the problem of the millers became one of the first cases to be investigated and prosecuted by the competition authority (CLICAC) in the mid-1990s. The authority brought suit before the courts against four companies in the country for collusion in fixing the price of wheat flour and in maintaining their respective market shares for wheat flour sales, for which they exchanged relevant information between November 1996 and September 1997. The case was recently decided in favour of CLICAC.29

Additionally, some countries erect entry barriers on flour imports by requiring compliance with specific regulations or norms. Such is the case of Resolution 94-2002 approved by the Council of Ministers for Economic Integration (COMIECO) in October 2002 for El Salvador, Guatemala, Honduras and Nicaragua, which accepted the technical norm for fortified wheat flour, Decree No. 30809, published

29 The sentence determined that on March 8, 1994 the millers had signed an agreement to set wheat prices and divide up the market estimated at that time at 120,000 quintales (1 quintal = 46 kg) per month. The defendants presented a high court appeal before the Supreme Court of Justice which was rejected as inadmissible; they have subsequently presented a second appeal which is awaiting disposition by the court (Fernández, 2005).

in Gazette No. 216 on November 8, 2002. This counteracted to a certain extent the effects of the liberalization of trade in wheat flour decreed by the Executive Committee of Economic Integration (CEIE) in October of the same year. Regardless, this latter measure represented an advance in the regional free trade of wheat flour reflected in expanded trading volumes for this product in Central America. Nicaragua, the country that most protected its milling industry with a 35% tariff until 2004, was one of the most favoured by this agreement. Nonetheless, there is a certain degree of differentiation in this product area, and even if imported flour is cheaper in Nicaragua, it is also of inferior quality, which has segmented the market. A significant price drop was by no means felt in all countries (Molina, 2007).

Given that the largest consumers of flour are the bread bakers, and the relatively high price of this input raises the price of this basic foodstuff in the region, some governments have removed import tariffs on flour from third-party countries such as Mexico. This has provided producers with a source of flour at competitive prices. Nicaragua's MIFIC agreed to liberalize flour imports by bakers and even extended to them a preferential line of credit for such imports. Nonetheless, in El Salvador as well as Nicaragua, it is difficult for bakers to import their inputs directly due to their small size and the lack of an established custom of taking advantage of external supply sources. As a result, the majority of their bakers continue to source flour in their domestic markets at monopolistic prices. One promising sign is that in July 2006 a new association of bakers in Nicaragua made its first direct import of flour.

2. The Air Transportation Market in Central America30

There has been little competition for passengers in the market for intra-Central American air transportation due to the extent to which the market is concentrated in the hands of Transportes Aéreos del Continente Americano (Air Transport of the American Continent; TACA) airlines since the early 1980s. This Salvadoran-owned company formed Grupo TACA, a business alliance that includes the participation of various Central America airlines, including Líneas Aéreas Costarricenses, S.A. (LACSA); Taca International Airlines, S.A.; Taca de Honduras, S.A. de C.V.; Aviateca, S.A.; and Nicaragüense de Aviación, S.A. (NICA). As such, practically all the region's national airlines belong to this conglomerate except Panama's Compania Panamena de Aviacion (COPA).

Midway through 2006, TACA was operating flights to 19 cities in Central America and the USA via its operations hub in San Salvador. TACA's principal route is North America – Central America, but its profit margins are particularly high on intra-regional flights. Outside of the region, the competition is greater and the company is not able to apply the same strategy.

30 This section is partly based on Molina (2007).

Central American airspace is protected through an intra-regional "open skies" agreement, but operators from outside the region are not accorded the same treatment. As a result, US and South American airlines are not permitted to operate flights between points within the region. Put another way, no airline other than TACA can pick up passengers in a Central American country and take them to another destination within Central America. In contrast, the market for North America – Central America routes is highly competitive.

Beginning in 1998, the regional market became somewhat more competitive due to an open-skies agreement signed between some Central American countries and the USA, allowing for the entry of new competitors into the market. Nonetheless, Grupo TACA continues to engage in monopolistic practices. For example, Panama's COPA airline requested and was granted, in December 2004, permission to operate additional flights on its routes from Central American countries to destinations in South America, Panama and the Caribbean. TACA filed several legal challenges to COPA operating a second daily flight on these routes, and on several occasions has succeeded in denying the Panamanian airline permission to land in El Salvador.31 Such anti-competitive conduct has had adverse effects across the region for consumers and airlines alike.

3. The Market for Carbonated Soft Drinks

Anti-competitive practices occur frequently in this market and have been challenged before competition commissions in Costa Rica and Mexico. However, these are practices that some multinational corporations, principally Coca Cola, engage in throughout many countries, including within the European Union.32 Such conduct consists of requiring both large and small-scale retailers, and even stadiums, amusement parks and other types of businesses agree to sell only Coca Cola brand beverages. This barrier also leads to the brand being sold at prices higher than those of alternative brands of carbonated soft drinks, as was the case when Big Cola began to compete with Coca Cola in Mexico.

In order to supply their product to retail resellers, Coca Cola frequently demands the exclusivity described above. This requirement may be applied in direct or indirect forms. For example, the Coca Cola Company or an affiliated bottler may provide the retailer with a small refrigerator in exchange for the commitment to stock it only with Coca Cola products. If the retail location is small enough that no other refrigerator of this type can fit, this guarantees that no other brand of cold beverages will be sold there. It is important to note that the small family stores that face these monopolistic practices are the channel for 75% of Coca Cola sales in Mexico, while

31 [online] skyscrapercity.com/archive/index.php/t-295767.html

32 Ibid.

restaurants, schools, clubs, hotels and entertainment facilities represent 24% of sales, and supermarkets only 1%.33

Consumption of this type of beverage in Mexico is enormous. By the end of the 1990s, 10% of the Coca Cola Company's global earnings was being generated in Mexico and, in 2004, this country was the sales location for 11% of the 19,800 million unit cases of soft drinks sold by the company. In addition, the Coca Cola Company regularly enters markets for related products seeking to replace domestic brands through large investments in advertising and promotions. This is the situation faced by firms producing fruit drinks, juices and nectars in Mexico. Coca Cola will use its Minute Maid brand to challenge Jumex, Jugos del Valle and the worker's co-operative Pascual Boing for control of this market.34 Special attention is needed to assure that the practices that shaped the carbonated soft drink market do not extend worldwide to that of juice, fruit drinks and nectars.35

Charges of monopoly practices in the sale of certain products were filed before competition officials in Costa Rica in 2001,36 and Mexico's FCC received several such complaints between 2000 and 2003. In both instances heavy fines were levied on the company accused of conducting these practices.

VI. Conclusions

There were no instruments for promoting market competition in Central American countries at the time when governments in the region carried out most of the economic deregulation process and privatizations of large public companies, which in many cases began to operate as private monopolies. Such monopolies and other market distortions that accumulated since the economic reforms posed a major challenge for the competition agencies once they were formed (in five of the seven countries studied, including Mexico, by the end of 2006), especially in the smaller economies. The countries that still need to create the legal and institutional framework for competition will face these problems perhaps to an even greater degree, but they also will be able to learn from the mistakes and successes of their neighbours.

Competition is now on the agenda of governments of most countries in the region. Competition policy has greatly expanded in recent years and between 2004 and 2006 not only have two new laws been passed (in El Salvador and Honduras),

33 [online] www.rel-uita.org/companias/coca-cola/cosecha-para-femsa.htm

34 [online] www.crain.com.mx/Snews/news_display.php?story_id=633

35 Ibid.

36 The suit in Costa Rica was against Coca Cola and Embotelladora Panamco Tica S.A. on the grounds that these companies engaged in relative monopolistic practices. In June 2001, the company Embotelladora Centroamericana S.A. brought suit against Embotelladora Panamco Tica S.A. for the same reasons. The latter company, one of the largest in Latin America, commonly imposes prices on the businesses that sell the products they distribute (Sittenfeld, 2006 and her chapter in this book).

but the pioneers in competition policy (Costa Rica, Panama and Mexico) have reformed or are in the process of reforming their laws, as we have previously noted. The trend is towards expanding the coverage or applicability of competition law by reducing the number of exceptions, incorporating the "clemency appeal" in order to have more channels for access to information on monopolistic practices and increasing fines to give them more dissuasive influence over potential transgressors. In addition to modifying the laws, these countries have also made efforts to improve training for judges and seek closer relationships and collaboration between competition agencies and the judicial systems. The countries that have recently passed their laws – El Salvador and Honduras – have already incorporated some of these improvements.

After at least ten years of experience in competition policy in three of the countries under study, we can see that the legislative model based on that of the USA, in which the courts play a major role in case resolution, does not appear to be the most adequate for countries such as those of Central America which lack expertise on both the judicial and competition-policy arenas. This is the situation in Panama, which even created special courts for cases that are often handled in a very protracted and unsatisfactory manner. It seems that a European-style system, in which cases could be handled through administrative channels, would be much more effective. Furthermore, there are numerous ways companies engaging in anti-competitive practices may avoid or delay the payment of fines, and such loopholes for legal manoeuvering must be eliminated to make competition laws more effective.

The lack of culture of competition is a prominent problem in Central American countries and a concerted effort to address this issue is indispensable. An effective application of competition law to anti-competitive practices, however, is an important means of legitimizing and promoting the knowledge and respect for the institutional framework needed for guaranteeing the healthy functioning of markets.

The relationship between regulation and competition is very important in this type of economy, and they should complement each other so that goods and services, especially of types previously provided by the state and now dominated by large corporations, can be provided in a technically adequate manner.

In small economies, the competition agency faces a special challenge in the case of market concentrations, given that these may be necessary in order to achieve important technological innovations and increased efficiency. A major challenge for the countries under study is how to distinguish such objectives from that of increasing profit margins through monopolistic practices. In this regard the competition agencies of Mexico, on the one hand, and Costa Rica and Panama, on the other, vary in their capacity to deal with this issue. In Mexico, the majority of cases are related to corporate mergers and takeovers, transactions for which participating companies must give regulatory officials prior notification. In the other two countries, where notification of merger or acquisition plans is voluntary, competition agency efforts are focused largely on other types of practices (generally horizontal monopolistic practices). El Salvador is the first SDC in Central America to require prior notification of mergers, and its competition commission will face challenges in carrying out its work, considering the scarce resources available to it.

As we have seen, competition policy has advanced considerably in recent years in the countries under study, but at the same time the further advances have been registered in competition policy at the international level, thereby posing new challenges for the region. For example, the local trend has been to apply the "rule of reason" instead of prohibiting certain practices per se, even when the practices are horizontal in nature. Even if this way of treating cases is undoubtedly more reflexive and probably leads to fairer resolutions, it poses the need to provide SDC with greater resources. It is also important to maintain a broad view of the changes experienced by these countries as a result of their insertion into global markets. They all have strengthened commercial ties through free trade agreements (FTAs) between each other or with third-party countries and this can open the door for large multinational corporations to take advantage by exerting their market power in these countries, or for other types of distortions to occur although improvements in competition also may take place in some markets as a result of these agreements. Hence, many FTAs include competition clauses.

In this context, regional and international cooperation on competition is crucial within Central America, and between the region and those countries with which it has strong economic ties such as the USA and Mexico. The possibility of collaborating on information sharing and coordinating actions to avoid anti-competitive practices that simultaneously affect various markets of the region (e.g. the cement, beverage, air transportation and banking industries, among others) would be of great help. This becomes even more significant when considering the limited resources available to all of the regions' competition agencies. It would also be very useful to make a great effort to harmonize competition rules at the regional level, create similar criteria for confidentiality and transparency of information and pursue cases of trans-border monopolistic practices. A seed has been sown for taking up this effort in the Guatemala Protocol of the General Treaty of Central American Economic Integration (1993), which includes the following stipulation: "In the trade sector, the states party to this agreement agree to adopt common dispositions to avoid monopolistic activities and to promote free competition in the countries of the region."

Bibliography

Adhikari, R. and J. Knight (2004), What type of competition policy and law should a developing country have? South Asia Economic Journal 5: 1–25.

Ansorena, C. (2007), Competencia y regulación en las telecomunicaciones: el caso de Nicaragua, (LC/MEX/L.785), serie Estudios y Perspectivas, No. 73, ECLAC/IDRC, July.

Avalos, M. (2006), "Condiciones generales de competencia: el caso de México", Serie Estudios y Perspectivas No 48, ECLAC/IDRC, Mexico, May.

Baumol, W.; J. Panzar and R. Willig (1982), Contestable Markets and the Theory of Market Structure, Harcourt Brace Javanovich, New York.

Cernat, L. and P. Holmes (2004), Introduction, competition, competitiveness and development: Lessons from developing countries, in P. Brusick, A. M. Álvarez, L. Cernat, and P. Holmes, United Nations Conference on Trade and Development (UNCTAD), New York and Geneva, June.

De León, I. (1998), Un enfoque alternativo de políticas para la promoción de la competencia en economías en transición, Boletín Latinoamericano de Competencia, N° 5 (third part), August, pp. 50–72 [online] (http://europa.eu.int/comm/competition/international/others/latin_america/boletin/boletin_5_3_es.pdf).

De León, I. (2000), The role of competition policy in the promotion of competitiveness and development in Latin America, Boletín Latinoamericano de Competencia. N° 11. December, pp. 57–72. [online] (http://europa.eu.int/comm/competition/international/others/latin_america/boletin/boletin_5_3_es.pdf).

De León, I. (2006), La promoción de la competencia en Centroamérica. Dificultades y posibilidades, in C. Schatan and M. Avalos (Coords.), Condiciones y políticas de competencia. Economías pequeñas de Centroamérica y el Caribe, FCE and ECLAC, Mexico D. F.

Evenett, S. (n.d.), "Competition policy in small developing countries: reflections on current literature", Power Point presentation.

Fernández, M. (2005), Condiciones generales de competencia en Panamá, Serie Estudios y Perspectivas, N° 37, ECLAC/IDRC, Mexico, September.

Gal, M. (2001), Site does matter: The effects of market site on optimal competition policy, Southern California Law Review, vol. 74, Los Angeles, CA.

Gal, M. (2003), Competition Policy for Small Market Economies, Harvard University Press, Cambridge, MA.

Gal, M. (2004), Competition: Prerequisites for development-oriented competition policy implementation, competitiveness and development: Lessons from developing countries, in Brusick (ed.), United Nations Conference on Trade and Development (UNCTAD), New York and Geneva.

Hernández, R. and C. Schatan (2006), Políticas de competencia y de regulación en el Istmo Centroamericano, in C. Schatan y M. Avalos (Coords), Condiciones y políticas de competencia. Economías pequeñas de Centroamérica y el Caribe, FCE y ECLAC, Mexico.

Khemani, R. S. (1997), Competition policy and economic development, Policy Options, October [online] (http://www.irpp.org/po/archive/oct97/khemani.pdf).

Khemani, R. S. and M. Dutz (1995), The instruments of competition policy and their relevance for economic development, in Frischtak (ed.), Regulatory Policies and Reform: A Comparative Perspective, (Pre-publication Edition), World Bank, Washington, DC.

Klemperer, P. (1995), Competition when consumers have switching cost: An overview with applications to industrial organization, macroeconomics, and international trade, Review of Economics Studies 62, 515–539.

Levy, S. (2005), Ensayos sobre desarrollo económico mexicano, Fondo de Cultura Económica, Mexico.

Levy, B. and P. Spiller (1996), Regulations, Institutions and Commitment, Comparative Studies of Telecommunications, Cambridge University Press.

Mariscal, J. and E. Rivera (2005a), New trends in the Latin American telecommunications market: Telefonica & Telmex, Telecommunications Policy, N° 29, pp. 757–777.

Mariscal, J. and E. Rivera (2005b), Organización, industrial y competencia en las telecomunicaciones en América Latina: las estrategias empresariales, serie Desarrollo Productivo N° 169 (LC/L.2423-P), ECLAC, Santiago, Chile.

Mariscal, J. and E. Rivera (2007), Regulación y competencia en las telecomunicaciones mexicanas, serie Estudios y Perspectivas N° 83, ECLAC/IDRC, June, Mexico.

Molina, F. (2007), Condiciones generales de competencia en países centroamericanos: el caso de El Salvador, Serie Estudios y Perspectivas N° 67, ECLAC/IDRC, Mexico, January.

Romero, A. and C. E. González (2006), Condiciones generales de competencia en Guatemala, serie Estudios y Perspectivas, N° 52, ECLAC/IDRC, Mexico, May.

Sánchez Ugarte, F. (2004), Análisis económico de la legislación federal de competencia económica en Competencia Económica F., Editorial Porrúa, Mexico.

Sánchez Ugarte, F.; P. García Alba Iduñate; J. Aguilar; F. Heftye; G. Moguer and S. Sarmiento (2004), La primera década de la Comisión Federal de Competencia, Comisión Federal de Competencia (CFC), Mexico.

Schatan, C. and M. Avalos (2006), Condiciones y políticas de competencia. Economías pequeñas de Centroamérica y el Caribe, FCE and ECLAC, Mexico.

Singh, A. (2004), Multilateral competition policy and economic development. A developing country perspective on the European community proposal, Series on Issues on Competition Law and Policy (UNCTAD/DITC/CPL/2003/10), Geneva.

Sittenfeld, P. (2006), La puesta en marcha de la agencia de competencia; algunas lecciones aprendidas del caso costarricense, PowerPoint presentation [online] (http://www.eclac.cl/mexico/competencia/index.htm).

Stewart, T. (2006), Evaluación de la relevancia de las leyes de competencia en países pequeños: el caso del Caricom, in C. Schatan and M. Avalos (Coords.), Condiciones y políticas de competencia en economías pequeñas de Centroamérica y el Caribe, FCE and ECLAC, Mexico.

Tábora, M. R. (2007), Condiciones generales de competencia en Honduras, serie Estudios y Perspectivas No. 82, ECLAC/IDRC, Mexico, May.

Yong, M. (2005), Condiciones generales de competencia en Costa Rica, ECLAC/IDRC, Mexico, unpublished.

This page intentionally left blank

Chapter 3
Models of Privatization and Development
of Competition in Telecommunications
in Central America and Mexico

Eugenio Rivera

Introduction

Telecommunications in Central America and Mexico have undergone unprecedented changes. The coverage attained in fixed telephony, and particularly in mobile telephony, was unimaginable just a few years ago. Despite this general trend, the countries of the region have taken varied paths in the privatization and liberalization of telecommunications, which have resulted in different levels of competition and performance in the industry.

The purpose of this chapter is to analyse, based on national studies on competition and regulation in telecommunications, how the different forms of privatization and liberalization of this industry or, conversely, the preservation of public enterprise have affected the intensity of competition in the industry, the coverage and the quality and price of services.

For this purpose, first we analyse the different forms of privatization and liberalization of telecommunications or their rejection, seeking to identify the causes leading to the application of the various models. Three courses of development can be pieced together in the region.

The first concerns the case of Costa Rica, in which the citizens rejected privatization and opted for keeping the Costa Rican Electricty Institute (ICE), also in charge of telecommunications, as a public enterprise. Honduras shows significant similarities with this course of development. The second involves the cases of Mexico, Nicaragua and Panama, where the strategy was to privatize and provide the enterprise with a period of exclusivity. This policy aimed at strengthening the enterprise prior to the liberalization of the market. The third approach was that adopted by El Salvador and Guatemala which was characterized by the simultaneous privatization and liberalization of the industry and also by the privatization of radio spectrum management.

Among the major options mentioned, privatization is considered by the literature an important factor for the sector's performance.1 Whether the monopolistic public enterprise was privatized as a vertically and/or horizontally integrated monopoly, or

1 On this subject see Rivera, 2006.

whether measures were taken so as to limit its monopolistic power through restructuring, or if certain precautionary steps were taken to limit its dominant position seems to have been crucial for the enterprise's behaviour and this will be looked into in this chapter. These different paths are the result of specific political configurations and coalitions and diverse types of historical performance by the public enterprise. The international context in which the different processes took place also had a strong influence. Thus, the Mexican reform can only be understood within the context of approval by the North American Free Trade Agreement (NAFTA), and the final destiny of ICE in Costa Rica will probably be determined by the Dominican Republic – Central America Free Trade Agreement (DR-CAFTA).

Together with the forms of privatization and liberalization, the industry's performance is associated with the existence of a legal framework favouring competition and effective conditions for its application. This latter aspect refers to the institutional capacity of both the sector regulator and the competition agency to effectively perform regulatory tasks and supervise competition, and their respective capabilities for coordination. It also refers to the ways in which the overall institutional framework operates, and particularly the administration of justice. The regulatory and competition agencies, in addition to undertaking tasks incumbent upon the Executive Branch such as administration of the radio spectrum and formulation of telecommunications and competition policies, also operate as administrative entities assisting in the administration of justice.2 Paradoxically, the Judiciary sometimes stands as a significant obstacle to the application of those policies, thus affecting competition and market efficiency.

In this context, Section I compares the different privatization and liberalization processes as well as their main determining factors. Section II analyses the main features of the regulatory and competition institutions, as well as their legal framework. The general business climate and its impact on a competitive atmosphere are also studied in this section. Section III studies the processes of opening up to competition, a crucial time for assessing the consistency of the regulatory framework and its application. Section IV focuses on the characteristics of the sector's industrial organization and the performance and development of competition in the industry's different segments. Section V analyses the evolution of rates in the countries of the region. Section VI identifies the trends influencing the future of telecommunications in the region and sets forth the main conclusions together with policy recommendations.

I. Privatization and Liberalization of Telecommunications in Central America and Mexico: The Diversity of Experiences

With the exception of Costa Rica, the seven countries covered by this study have furthered forms of privatization and deregulation of telecommunications. However, these forms of privatization and liberalization of the sector have varied.

2 We wish to thank Gustavo Paredes, while being a Commissioner of CLICAC, for having called our attention to this link.

1. The Case of Costa Rica

Costa Rica has a long experience in open trade practices and in the establishment of free trade agreements. Recently it took part in the negotiations leading to the signing of DR-CAFTA, although it is yet to be ratified by the Legislative Assembly. Even though the main political forces have expressed their support for this agreement, citizens' rejection persuaded the President of the Republic to postpone seeking its ratification by the said Assembly. The chief reason for this rejection is the commitments assumed by Costa Rica in telecommunications in Annex I.

In DR-CAFTA, Costa Rica pledges to promote the opening up of telecommunications services gradually and selectively, and strictly in accordance with the social objectives of universality and solidarity in the provision of such services. To that end, Costa Rica would have had to enact, by December 31, 2004, at the latest, a legal framework that would allow the strengthening and modernization of ICE.3 Likewise, by virtue of the Agreement it would have to allow other providers to supply telecommunications services under no less favourable terms and conditions than those established by, or granted in accordance with, its legislation in force to January 27, 2003. Costa Rica would not have permitted other telecommunications services providers (including international) on a non-discriminatory basis to compete in order to supply directly to the client, with the technology of its choice, the following telecommunications services: (i) private network services, at the latest by January 1, 2006; (ii) Internet services, at the latest by January 1, 2006; and (iii) mobile wireless services, at the latest by January 1, 2007. Finally, the country would have to apply the new regulatory framework for telecommunications services as of January 1, 2006, in accordance with the following provisions: universal service, independence of the regulatory authority, transparency, allocation and utilization of scarce resources, regulated interconnection, access to, and use of, information services supply networks, competition, undersea cables systems and flexibility in technology options.

The DR-CAFTA negotiations were accompanied by intense public debate and social conflict. The debate continued after its signing in January 2004. The situation was further complicated when proceedings were instituted against two ex-presidents involved in cases of corruption in connection with the Social Security Fund and ICE, probably the two leading institutions of the so-called Costa Rican model. These cases appeared to confirm all citizens suspicious who viewed the different initiatives for modernization of the public electricity and telecommunications enterprise only as attempts by privileged groups to appropriate, at a low price, the national wealth.4 The early start of the campaign to elect a new president in

3 In February 2006 presidential election, the candidate of the Citizen's Action Party actively opposed this Free Trade Agreement. Nevertheless, the winning candidate, Oscar Arias, supported such agreement all along.

4 For an analysis of the different proposals for reform of ICE, see Rivera (2006).

February 2006 led the authorities to postpone the parliamentary debate on the Agreement until after the elections.

Costa Rica is the only country among those studied5 that has rejected the privatization and liberalization of telecommunications. The main reasons for this stance are as follows. (i) The public electricity and telecommunications enterprise has been associated with the most positive aspects of the Costa Rican model, which enabled a relatively poor Central American country to emerge as the most developed in the region. (ii) Associated with the above, ICE took electricity and telecommunications services to every corner of the country and achieved higher levels of coverage than most Latin American countries. While in the rest of the region users had to wait years for access to a telephone, Costa Ricans were able to gain access so much more quickly. (iii) ICE has also played a crucial role in helping the country reach levels of competitiveness that have enabled it to participate favourably in the most modern international chains of production. National Telecommunications Institute's (INTEL) investments are one of the main examples of this situation. (iv) After the privatization and liberalization of telecommunications in the rest of Central America and Mexico, fixed telephony in Costa Rica remained at the head of the industry in terms of coverage, and was placed second in Latin America.6

2. Privatization with Exclusivity: The Cases of Mexico, Nicaragua and Panama

In the cases of Mexico, Nicaragua and Panama, the privatization process was accompanied by the right of an exclusivity period for the telecommunications enterprise, prior to the opening up of the market to competition. The similarities among these experiences, however, conceal significant differences in telecommunications public policy.

Mexico's case is undoubtedly a paradigmatic example of nationalism in the process of privatization and opening up of telecommunications. The enterprise was privatized as a national, vertically integrated monopoly and was granted a six-year exclusivity period. The new General Law of Telecommunications (LGT) was passed more than five years after privatization and the regulatory body was established shortly before the end of the exclusivity period. The privatization of telecommunications in Mexico formed part of a broader process of reform begun in 1986 with the accession to General Agreement on Tariffs and Trade (GATT) and the government's decision to establish a free trade agreement with the USA and Canada. Despite opposition from large social groups, the hope that this policy would contribute to reactivating the economy and to its

5 The group of countries includes Costa Rica, El Salvador, Guatemala, Honduras, Mexico, Nicaragua and Panama.

6 The only other country in Latin America that has rejected privatization of its telecommunications enterprise has been Uruguay.

long-term growth succeeded in rallying majority support in society. The poor performance of the public telecommunications enterprise constituted a forceful argument in favour of the reform and, at the same time, privatization became a fundamental component of the modernization of the economy (Mariscal and Rivera, 2005a).

Telecommunications policy in Mexico cannot be understood unless it is inserted into the context defined by the decision of the administration of Salinas de Gortari (1988–1994) to reorient Mexico's development from the import substitution model towards an export model based mainly on close association with the USA and Canada. Betting on NAFTA led the Mexican government to use the privatization of Teléfonos de México (TELMEX) as evidence of its commitment to approval of the Agreement. With the privatization of the enterprise the Salinas administration wanted to show its split with the national revolutionary tradition and its willingness to further private sector-driven capitalism. Government policy was also determined to privatize because of its fear that the telecommunications enterprise would suffer a rapid process of denationalization or become a victim of competition from major foreign operators, particularly from the USA. In this context, the enterprise was strengthened financially as much as possible so as to become attractive to domestic investors (Mariscal and Rivera, 2007). The decision was taken, moreover, to privatize the enterprise as a vertically and horizontally integrated monopoly and it was decided that the dominant share among the investors should be Mexican.7 The exclusivity period paved the way for an important set of conditions for the modernization of the enterprise, digitalization of its networks and financial strengthening which stood it in good stead to face the generalized opening up of the industry to competition as of 1997. Together with tax benefits, the tariff policy set in the Concession License was highly favourable to the enterprise, allowing high rates of return, which meant that in 1999 Mexico registered the Organisation for Economic Co-operation and Development's (OECD) highest prices for telecommunications services (OECD, 1999). This policy enabled the company to begin its international expansion in 1998 and since then it has become, together with Telefónica España, one of the two dominant companies in the region.

If in Mexico's case the exclusivity period and other policies applied in favour of TELMEX can be explained by the aim of contributing to shape a national champion capable of vying with US companies for the position of privilege within Mexican borders, in the cases of Panama and Nicaragua this policy is harder to understand.

In Panama, on July 14, 1992, the Legislative Assembly, controlled at the time by the Christian Democrat Party, succeeded in passing a law "whereby the privatization process for state enterprises and services is established and regulated." However, the opposition introduced an Article which indicated that the law would not be applicable to the privatization of public utility enterprises, INTEL, Institute of Water Resources and Electrification (IRHE) and National Aqueducts and Sewerage

7 The concession licence obliges TELMEX to maintain its capital structure and Board of Directors, so that the power to determine the company's administrative control and management should fall chiefly on the Mexican partners (Mariscal and Rivera, 2007).

Systems Institute (IDAAN), or to the services they provide, and that a special law to approve their privatization would be required. Paradoxically, one of the parties that authored the above-mentioned Article took office in government in 1994 and proposed a bill, passed in 1995, paving the way for the privatization of INTEL.8 The immediate causes of privatization appear to have been problems of supply in fixed telephony, the non-existence of mobile services and the fiscal problems of the Panamanian State, which made it difficult for the state to carry out the necessary investments in this sector. (González, 2007).

The opening of envelopes containing the companies' offers took place on May 20, 1997. The company Cable & Wireless won the bid for the enterprise by offering $652 million as against General Telephone and Electronics' (GTE) $451 million. The benchmark price had been established at $500 million, so $152 million above that price was obtained. With regard to mobile telephony, in 1996 an invitation for tenders was issued for Band A for the provision of mobile services throughout the country. The winning bid was from Bellsouth, in alliance with Multiholding,9 which paid $72.6 million. Motorola, for its part, offered only $42 million. Band B was assigned to the winner of the bid for INTEL (which was Cable & Wireless, as mentioned), which would have to pay the same amount paid for Band A (González, 2007).

What are the reasons behind the Panamanian government's decision to grant an exclusivity period in fixed telephony, award a mobile licence as part of the sale of the fixed telephony enterprise and limit mobile telephony service to two operators? Three explanations have been put forward in the debate on the matter: (i) the presence of natural monopoly in fixed telephony; (ii) the government's objective of maximizing its revenues as owner of 49% of the shares led it to turn over the sector in exclusivity – The present value of cash flows is greater under that system; (iii) radio spectrum restrictions did not make the entry of more operators viable (González, 2007).

The first argument may be convincing to justify the sale of the enterprise as a monopolistic operator. It is not, however, the grounds for a period of exclusivity. Exclusivity is justified precisely as a period in which the enterprise renews itself and reaches a critical threshold that makes it possible for it to face the competition. But the characteristics of a natural monopoly rule out any attempt at entry of new firms, and therefore additional barriers seem unnecessary. The second argument is undoubtedly the relevant one. The concession of exclusive rights allowed the government to maximize its revenues at the time of sale and, as owners of 49% of the shares, profit from a greater flow of dividends. The fiscal reasons do not, however, seem sufficient to explain the continued public presence in the ownership of the enterprise. In this regard, the Democratic Revolutionary Party (PRD), the dominant group in the coalition that assumed the Presidency of the Republic in 1994 and was characterized by strong nationalistic tendencies, saw in public participation in the

8 Ernesto Pérez Balladares, of the United People Alliance, headed by the PRD, obtained 355,307 votes of a total of 1,066,844 valid votes. The PRD is the party founded by the revolutionary national leader Torrijos.

9 The shareholding was Bellsouth USA 43.7% and Multiholding Corporation 56.3%.

company INTEL a mechanism to ensure consistency between economic growth objectives and modernization of the enterprise and national interests. The Panamanian government's interest is reflected by the fact that eight years after 51% of the enterprise's shares were privatized, the Panamanian State continued to hold 49% of the shares. A further expression of this situation is the existence of ongoing negotiations between the government and the enterprise to obtain lower prices for public institutions than those paid by other users of telecommunications services (González, 2007). In this regard, Panamanian policy in privatization and liberalization of telecommunications may be interpreted as a version of the policy of national champions in the case of small countries.

Nicaragua's case also has its peculiarities. It is the region's poorest country, with long-standing high external indebtedness resulting from having to finance its current account deficits and a substantial percentage of its fiscal deficits with international aid or concessionary loans. This situation makes the country highly vulnerable to international pressures in favour of privatization. Privatization of the public telecommunications monopoly was put forward as a solution to the problem of demands for indemnity on properties confiscated by the Sandinista government. The funds obtained would be allocated to finance the bonds issued in favour of persons with unfairly confiscated properties. The proposal thus succeeded in building political consensus on privatization of the industry (Ansorena, 2007).

The privatization process was preceded by a significant investment effort. Between 1992 and 1994, Nicaragua achieved the highest rate of investment and modernization of telecommunications in the whole of Central America, so that the assets of the public operator were the most modern in the region at the end of that period. Investments during that time span totalled $103.9 million, 72% of which was earmarked for the acquisition of infrastructure and equipment. As in the two preceding cases, a horizontally and vertically integrated monopoly was privatized, with an exclusivity clause for a period of three years. The exclusive concession also included national and international long distance (LD) and the option to acquire a mobile telephony licence. In its part, the winning company had to pledge to meet network expansion goals, improvement of service quality and creation of a technological platform (essential equipment) to allow interconnection and thus increase the participation of other operators once the exclusivity period was over. The privatization model included the sale of 40% of the shares, with majority voting power equivalent to 57%; employees could have access to 11% and the remainder 49% would remain as property of the state for subsequent sale on the domestic or international stock market. A number of difficulties postponed the privatization until 2001.10 Finally, on December 12, 2001, after a public bidding process, 40% of the

10 Political difficulties and strong public opposition to the privatization of the only profitable public services enterprise postponed approval of the sale by the National Assembly until 1998. The first competitive bidding process begun in May 1999 was declared void when the two interested companies, TELMEX of Mexico and Telefónica of Spain, decided to withdraw. The second attempt, six months later, had the same outcome, when a consortium made up of MCI WorldCom, TELMEX and France Telecom withdrew from the process (Ansorena, 2007).

enterprise's shares were turned over to the Swedish-Honduran consortium Telia Swedtel Ab – MEGATEL EMCE for a period of 20 years. The price paid was $83 million, far below the benchmark price of $203 million (Ansorena, 2007). In 2004, in contrast to the above case, the government proceeded to sell the remaining shares, awarding them without a bidding process to the Mexican company América Móvil, part of Grupo Carso. In parallel, the same company acquired MEGATEL of Honduras, as a result of which América Móvil became the sole shareholder of Empresa Nicaraguense de Telecomunicaciones, S.A. (ENITEL). Furthermore, with the purchase of MEGATEL, América Móvil became an operator with two cellular telephony licences, one from ENITEL and the other from Servicios de Comunicaciones Personales Inalambrica (PCS SERCOM).

If in Panama's case the exclusivity policy expressed an interest in maintaining a public presence in the sector, in Nicaragua's case it represented only a concession to the investor. Protecting an international telecommunications operator does not appear to make sense from the point of view of the public interest. On the other hand, the sale of the enterprise to the Mexican operator, which enabled the latter to strengthen its position in the fixed telephony segment and also in national and international LD and, at the same time, made it easier for that same company to have majority control of mobile telephony, significantly reducing the intensity of competition, is not easy to explain. The severe political problems affecting Nicaragua between 2003 and 2006, the consequence of which has been, among others, the parallel and conflictive existence of two regulatory bodies, may contribute to understanding that situation. We will return to this topic further ahead.

3. Privatization without Exclusivity: The Cases of El Salvador and Guatemala

In the cases of Guatemala and El Salvador, privatization and opening up to competition took place simultaneously.

In El Salvador there appear to be two driving forces behind the privatization and reform of telecommunications. First, the decision of the 1989–1994 Administration to promote a programme of comprehensive reforms, drawn up with the support of a working group headed by Arnold Harberger and aimed at building a new economic model. This programme included initiatives to stabilize macroeconomic imbalances (inflation, fiscal deficit, current account, competitive exchange rate), a structural adjustment programme (privatization of the banks, liberalization of foreign trade and impetus for tax reform) and additional policies aimed at propelling an export-led economic growth (Argumedo, 2007). The second was the deep discontent of the main entrepreneurial groups with the situation of telecommunications.11

11 The export strategy was facing serious difficulties due to lack of infrastructure. Textile maquila, for example, needed modern telecommunications to participate in international "just in time"

In this context, as in the case of Nicaragua, the Administration decided to promote a wide-ranging programme of investments and a policy expressly designed to raise the enterprise's productivity. Thus, between 1990 and 1993, the enterprise called Administración Nacional de Telecomunicaciones (ANTEL) invested $133.1 million, i.e. an average of 0.6% of gross domestic product (GDP) annually. The number of lines was increased from 2.9 per 100 inhabitants in 1989 to 5.5 in 1993; the number of employees was reduced from 48 per 1,000 lines to 22 in 2003; and the institution's profits improved (Argumedo, 2007). The following Administration (1994–1999) decided to intensify the progress made, creating the Presidential Commission for the Modernization of the Public Sector (CPMSP) with the aim of decentralizing or privatizing public services such as electricity, telecommunications, water, public transport and ports, as well as analysing the reform of the Pensions System. As in other cases, the reform of telecommunications was assigned an exemplary role, which led to its being carefully prepared. Indeed, between 1994 and 1995, various studies were commissioned which had to include a diagnosis of the problems and the requirements for creating a competitive market in telecommunications. In 1996, the new Law of the sector was passed and the regulatory body was created. In 1997, the enterprise was restructured in such a way that a fixed telephony company was created and another one also to take charge of mobile operations. Finally, in 1998 both companies were privatized. France Telecom acquired 51% of the shares of Commonwealth Telephone Enterprises (CTE), S.A. de C.V., the fixed telephony company, for $275 million (the benchmark auction price was $268.8 million), and Telefónica España bought INTEL, the mobile company, for $41.0 million (the benchmark auction price was $11.9 million; Argumedo, 2007).

In Guatemala, privatizations began in the early 1990s. In 1993, a selective process to sell low-utilization and low-performance state assets began with the aim of financing public-benefit programmes. The most distinctive feature of economic reform until 1996 was the sale of state assets to finance the fiscal gap and obtain resources for activities and programmes on the social agenda. Most of the privatizations carried out in this period (mainly the sale of the national airline Empresa Guatemalteca de Aviacion, S.A. (AVIATECA) and the concession awarded to Comunicaciones Celulares (COMCEL) for Band A) were severely criticized for the

production strategies. In the early 1990s, the waiting time to obtain a new telephone line was more than two years, there were problems of quality in communications, the geographical coverage of the network was very limited, and the damage caused by the civil war was to demand resources to replace the infrastructure. This situation limited the State's possibilities to attract foreign investment. The high rates in long distance telephony, which subsidized local calls, hindered the competitiveness of exporting companies. A survey on obstacles to competitiveness carried out by the World Bank showed that over 58% of the companies interviewed stated that they had problems with the service, 71% of these said they faced the problem of overloaded lines, 14.5% pointed out crossed lines and interruption and more than 78% stated there were not enough lines to be able to acquire new ones. Other indicators showed low-quality service since 75% of faults took between four and 20 days to repair, whereas the rest took even longer. Around 35% of calls could not be completed due to high congestion (Argumedo, 2007).

procedures followed, which lacked transparency and left the population with a bad aftertaste regarding privatization processes. Towards 1996, after the signing of the Peace Accords, the political regime was becoming consolidated and although the growth rate was satisfactory with respect to the region, the country was facing a fiscal situation that threatened macroeconomic stability. In this context, a new government took office, headed by President Álvaro Arzú. Progress in the modernization of the state as a means to build a more efficient and competitive economy was the prime objective of the Government Program for the period 1996–2000. Special importance was placed on physical infrastructure, particularly in telecommunications (Urízar, 2007).

The government analysed three options: (i) keeping Empresa Guatemalteca de Telecomunicaciones (GUATEL) as a public enterprise and promoting initiatives aimed at strengthening it; (ii) selling GUATEL with a five-year period of exclusive concession; and (iii) selling GUATEL, thereby immediately opening up the market to competition and concentrating the government's activity on its regulatory role. The first option was discarded, as it did not adjust to the principles of subsidiarity promoted by the government. The second option proved attractive, particularly from the standpoint of public finances. However, the five-year concession of a monopoly was considered a risk, since a benefit of that nature would lead the private investor to do everything possible not to lose it. Thus, the reform would lead to nothing more than the passage from a public monopoly to a private one. To avoid that risk, the second option was also discarded. Despite the fiscal costs involved, the government opted for the third possibility. It was believed that a competitive atmosphere would contribute to attain more quickly a modern telecommunications services and at low prices (Urízar, 2007).

The government rejected the possibility of dividing the enterprise and preferred to sell it as a vertically and horizontally integrated operator. It was considered that its destructuring could hinder attracting investors and further affect the sale price. Efforts were made to neutralize the impact of this decision on competition through the approval of a Telecommunications Law. Political and legal difficulties of various kinds made it necessary to create a new enterprise, Telecomunicaciones de Guatemala, S.A. (TELGUA), to which all the assets and liabilities of GUATEL were transferred, leaving the latter with a minimal structure to enable it to continue serving community telephony. TELGUA took over the provision of local and LD telephone service and entered the market of the Personal Communications System (PCS). On October 1, 1998, the sale of 95% of TELGUA's shares to the company Luca, S.A. was finally concluded. The process was audited by Arthur Andersen and the value of the transaction was $977 million. Subsequently the owners of TELGUA established a partnership with TELMEX, an international operator, which acquired 45% of the shares. Prior to the Telecommunications Law, frequencies were obtained by authorization from the Executive Agency by means of concessions or authorizations. Chapter II (Articles 54–56) of the Telecommunications Law created the legal concept of "Frequency Usufruct Titles" (TUF), giving the spectrum the nature of an economic good rather than a public good. Thus, the category of regulated (free) Bands, which are allocated as TUF and can be freely transferred and utilized, within certain technical limitations, was created. TUFs detail the hours and area of operation, the maximum transmission power and the maximum interference permitted within the area of coverage (Urízar, 2007).

4. The "Honduran Way"

In the case of Honduras, with the enactment of the Law on Telecommunications (LMT) on October 31, 1995, Empresa Hondurena de Telecomunicaciones (HONDUTEL) became the responsibility of the Ministry of Public Works and Transport and later12 this responsibility was conveyed to the President of the Republic. Subsequently, these tasks were entrusted to the recently founded National Telecommunications Commission (CONATEL). Finally, the development and operation of telecommunications services were assigned to HONDUTEL and private companies. One fundamental feature of the LMT was the reform of the Organic Law of HONDUTEL, permitting association with private investors.

In this context the "capitalization process" arose for the purpose of attracting foreign private capital. At the same time, the enterprise was granted a concession in the following terms: "[F]or 10 years following the date of entry into effect of the Framework Law for the Telecommunications Sector, HONDUTEL shall provide, in an exclusive manner, national and international telephony services and telegraphy services in places where there is no other means of communication in the country, and such exclusivity shall be incumbent upon the company or companies established by HONDUTEL" (quoted by Tábora, 2007). In October 1998, the National Congress of the Republic approved Decree 244-98 which broadened the concession to 25 years and extended it to the non-exclusive operation of the carrier service. To modernize the sector, a mechanism was promoted that consisted in capitalizing a new enterprise with the assets of HONDUTEL. The new company, Compania Hondurena de Telefonos, S.A. (HONDUCOM), would be capitalized by means of a public bidding process in which 47% of the shares would be transferred to a technical partner who would take over the management of the enterprise, 2% would be sold to the workers and the state would keep 51% through HONDUTEL.

On signing a letter of intent with the International Monetary Fund (IMF) in 1997, a new distribution of shares was agreed upon, by which the state would keep 47% of the shares, 2% would be sold to the workers and 51% would be assigned to the strategic partner. A number of problems, among them Hurricane Mitch, delayed the process.

Finally, in 2000, seven companies showed an interest in acquiring the above-mentioned block of shares. However, only one of them, TELMEX, made an offer. The offer was only just over one third of the benchmark price ($300 million), as a result of which on October 16, 2000, the bid was declared void and in May 2001 it

12 On October 25, 1977, by virtue of Decree 118–97.

was decided to suspend the process until another arrangement could be found (Tábora, 2007).

The difficulties in capitalizing the enterprise led to the design of the programme "Telefonía para todos" (Telephony for all – TpT) with the aim of benefiting users and the national economy by addressing unsatisfied demand for telephone service and improving service quality and attention to customers, thus laying the foundations for the opening of the market as of 2005. To that end, competition would be introduced in fixed voice services, private participation in the sector and the supply of new services would be encouraged, thus stepping up the growth and modernization of the sector.13

The TpT programme envisaged broad access to the telecommunications market. In fact, all interested parties could sign a contract and receive equal treatment from HONDUTEL. Specifically, HONDUTEL should sign a contract with any party interested in establishing itself as a marketer of sub-operator type that complied with the following requirements: (i) being legally incorporated as a business enterprise in accordance with the country's laws; (ii) in the case of natural persons, having the legal capacity to contract and in the case of foreign companies, being authorized to carry out business transactions in Honduras; and (iii) submitting a sworn statement to the effect that neither the applicant nor any of the partners holding more than 10% of the capital stock are included in the cases provided for in Article 92 (Sections e to k of the General Regulations of the Framework Law for the Telecommunications Sector). Having complied with the above requirements, the interested parties would proceed to sign the Marketing Contract, after which they would obtain their registration and the licences for use of the radio spectrum (if necessary). Likewise, they would be assigned blocks of numbers and would have to pay the established rates.

The TpT programme introduced two important innovations. The first stems from the creation of the concept marketeer sub-operator, who is defined as "the one that receives from HONDUTEL an extension of the rights granted by Law, to enable it to directly provide the public telecommunications services authorized to HONDUTEL. To provide the services the sub-operators may make investments in infrastructure in order to expand the public telecommunications network, contributing both facilities and value-added services." Even more important is the implicit annulment of HONDUTEL's exclusivity (Tábora, 2007).14 Finally, the TpT decree

13 The programme "Telephony for All – Modernity for Honduras" (TpT) was approved by means of Executive Decree No. PCM 018–2003 dated September 23, 2003, and ratified by Legislative Decree No. 159–2003, published in the Official Gazette on October 24, 2003 (Tábora, 2007).

14 In this regard, Tábora points out that "On the other hand, the same TpT Decree, with the aim of permitting anyone wishing to invest in telecommunications to do so, eliminated all kinds of restrictions for those wanting to enter the market. This situation, due to the principle of egalitarian treatment and non-discrimination established by the LMT, left in contradiction and legal inconsistency all the constraints included in this matter in the concession contract of the mobile operator, CONATEL being obliged to amend the Concession Contract in these aspects, a situation which gave rise to dissent among the members of CONATEL, and which was interpreted by some as an example of inconsistency in sectoral policy on the part of the government, and as an irregular situation by others, even though the amendment was executed by a motion of the National Congress and was duly approved by Congress once it was made effective." (Tábora 2007).

ended segmentation in the provision of telecommunications services, since the sub-operator can use the same infrastructure to provide the different services, which also promotes technological convergence.

The impact of the "Honduran way" toward the liberalization and modernization of telecommunications seems promising. By September 2005, 40 companies had obtained registration as sub-operator15 of which 17 had started commercial operations and installed 63,000 telephone lines,16 two thirds of which belong to MULTIFON (Tábora, 2007).

II. Regulation and Competition in Telecommunications: Institutional Problems

1. Introductory Considerations

In the region's experience, with the exception of Costa Rica, public telecommunications enterprises traditionally have been unable to provide adequate services. Hence, privatization emerged as the possibility to overcome these problems by incorporating private management capacity into service provision.

However, the fact that telecommunications companies were natural monopolies made it necessary to create a regulatory agency to ensure consistency between private operation and public interests. International evidence was conclusive to the effect that a good law was not sufficient and that a regulatory agency was needed with the necessary resources to effectively apply the regulatory framework.

The problem is that the regulatory agencies created in the region began to operate either long after the privatization process (Mexico's case) or simultaneously with the privatization and the opening up to competition. Thus, these agencies are weak and face serious problems associated with asymmetry of information. Moreover, the regulatory agencies of small countries have to face transnational corporations which have major financial and technical backing. The technical capture is, therefore, a permanent possibility that prevents, or at least seriously hinders, effective regulation. The difficulties regulators face to obtain the required information seriously hinder their task and hence the need to promote competition as the main mechanism that can assure the industry of an efficient performance. However, the types of privatization and opening up to competition in the region have not been so favourable to the development of competition. Although the situation varies from one segment and from one country to another, the fact is that companies other than the incumbent effectively operating in the market are not sufficient in number or

15 The TpT Decree establishes that once HONDUTEL's exclusivity period is over, sub-operators may continue offering the services, whether under the same scheme of sub-operator or directly requesting their empowering title accrediting them as operators (Tábora, 2007).

16 Equivalent to one sixth of the total lines installed (Tábora, 2007).

in size to create the necessary competitive climate for the telecommunications market to operate efficiently.

In this context, the competition agency should play a crucial role in eliminating obstacles to entry and preventing anti-competitive conduct. Despite the progress made,17 the existing competition agencies in the region show serious difficulties in efficiently protecting competition in the telecommunications sector. In some cases the general law on telecommunications assigns the regulatory agency tasks designed to protect competition in the sector, but these are very general standards that are frequently almost inoperative and that definitely do not represent an effective means of promoting a competitive market.

In this context, the aim of this section is to identify the main problems regarding design of the regulatory framework for telecommunications and the institutional system for regulation and promotion of competition in this market.

2. Institutional Design and Competition18

Academic reflection and analysis of international experience regarding privatizations and the transition towards a competitive market in telecommunications tend to agree that the success of the process is associated with it being effectively planned; being preceded by the enactment of a Telecommunications Law that defines the rules of the game; and with the creation of a regulatory agency with the necessary authority and resources to efficiently apply the regulations. This institutional framework should be complemented with the prior establishment of the competition agency (Rivera, 2006). As we shall see, the experiences in the region are still far from this model.19

Although on different dates, Panama, Nicaragua, Guatemala, Honduras and El Salvador approved new Telecommunications Laws and created specialized regulatory bodies prior to the privatizations process. Up to the end of 2005 only three countries of the region had competition authorities: Costa Rica, Mexico and Panama. In Costa Rica's case, however, the issue is not important in view of the existence of the public monopoly in telecommunications. Furthermore, the Competition Law explicitly denies the Commission on Promotion of Competition (COPROCOM) any power regarding the sector. Further ahead we will return to the cases of Mexico and Panama after dealing with the cases of the countries that have lacked a competition authority.

17 The Superintendence of Competition of El Salvador began its operations on January 1, 2006, and Honduras was due to set up its agency six months after the Law was passed (early 2006).

18 For a detailed analysis of the Central American telecommunications regulatory framework, see Rivera (2006).

19 Costa Rica is once again an exception, since it lacks a specialized regulatory body. The state telecommunications and electric power enterprise still maintains important regulatory powers, particularly as regards spectrum management.

In El Salvador the Law creating the General Superintendence of Electricity and Telecommunications (SIGET) was passed by Legislative Decree No. 808 of September 12, 1996. The competition-related problems that could be faced by a telecommunications industry, particularly in a small country, led different sectors, as of the early 1990s, to put forward legal proposals to punish abuse of market power by one company to the detriment of its competitors. These proposals, however, were rejected, and therefore Article 111 of the Telecommunications Law granted SIGET the authority to prohibit anti-competitive practices. Unfortunately the law does not establish mechanisms for their study and sanction (Argumedo, 2007).

SIGET's highest authority is the board of directors, which is made up of one director appointed by the President of the Republic, who performs the duties of superintendent, one director elected by private sector trade associations legally established in the country (through the National Private Enterprise Association – ANEP) and one director appointed by the Supreme Court of Justice (CSJ). The directors are designated for a period of seven years in office and may be nominated again.20 However, in the first nine years of operation SIGET had four superintendents (average tenure two years and one quarter), which reflects the strong tensions to which the entity has been subject. The case of El Salvador is also noteworthy due to the presence, in SIGET, of a director elected by private sector trade associations, in particular the above-mentioned ANEP.

In general, in Latin America, there is much discussion regarding the possibility of the governments exerting pressure on the telecommunications regulatory body to favour short-term interests. Instead, in the Salvadoran case concern arises over the regulatory body being subject to the influence of the private sector. One would think that SIGET's director could represent the interests of the companies using telecommunications services and therefore that conflicts of interest do not exist. However, the public interest represented by SIGET could at times be inconsistent with the interests of business chambers.

The Competition Law was passed in late 2004 and entered into force on January 1, 2006, with the creation of the Competition Superintendence. One of the main challenges would be its coordination with SIGET.

In Guatemala's case, regulation of telecommunications is the responsibility of the Superintendence of Telecommunications (SIT). When the law creating the agency was under discussion, the possibility of granting the entity full autonomy was suggested. However, the support of two thirds of Congress could not be garnered for that purpose. It was therefore decided to make SIT dependent on the Ministry of Communications, Infrastructure and Housing (MICIVI), which, in turn, was to appoint and could remove the superintendent. This official's authority depends on the minister of MICIVI (Urízar, 2007). Subsequent amendments to the Law further undermined the authority of the Superintendent of Telecommunications,

20 A reform approved in December 1997, established the director from the private sector and the one from the CSJ could remain in office for five and three years, respectively.

first (by virtue of Decree 47–2002 Articles 13 and 15 were amended and 14 was eliminated), through a reduction in his financial resources and by conditioning his budget to a government decision; and second, by stripping him of the power to proceed judicially and administratively against transgressors of the law (modifications of Decree 15–2003 which amended numerals 2 and 3 of Article 81, relative to infringements and fines; Urízar, 2007).

In Honduras, the LMT Sector, enacted through Decree No. 185–95 of October 31, 1995, created CONATEL to regulate and oversee the development and operation of telecommunications. CONATEL has the authority, among others, to collaborate with the President of the Republic to formulate telecommunications policies and ensure their effective execution by means of regulation and coordination actions. It therefore assumes the duties entrusted in other countries to the Ministry of Communications. CONATEL's Commissioners are appointed by the President of the Republic, may remain in their posts for up to four years, and may be nominated for additional periods. Their terms of office coincide with that of the President of the Republic and they may be removed whenever the authority deems it appropriate.

The LMT establishes that it is the responsibility of CONATEL to promote competition in the telecommunications services. In this regard CONATEL must ensure that operators of telecommunications networks provide access, on equal conditions, to other operators and users in the same or similar circumstances. Nevertheless, despite the fact that these duties were entrusted to CONATEL, the LMT, on establishing the breaches to be sanctioned as serious, makes no reference to infractions against free competition (Tábora, 2007). The type of privatization process in Honduras, already described, which maintains the incumbent enterprise as public, has so far minimized the problems that this legal and institutional design could pose to the industry's development.

In 2005, the debate ended and the Competition Law was passed. The new legislation has pre-eminence over sector legislation, but promotes coordination with the sectoral regulatory body, especially when the markets are reasonably open to competition as in the case of telecommunications. The reasons that led to demanding coordination are the following: (i) the need to have technical knowledge of subjects related to the sector during the transition to, and consolidation of, the new competition agency; (ii) the need to define the setting of competition in advance and not be limited to applying sanctions retrospectively for conduct contrary to competition or to restructuring the sector; and (iii) the need to apply policies unrelated to competition but which the government considers important, for example, on universal service, national security and protection of users.

Finally, as in the other countries without a competition authority, in Nicaragua the Telecommunications Law provides for a competitive mechanism in the provision of services. However, Article 24 of that Law envisages the possibility of granting concessions with exclusivity for a period to an operator. The Law prohibits anticompetitive practices and establishes that operators are obliged to provide satisfactory access at competitive rates to the telephone network of service providers whose licences have been authorized by Instituto Nicaragüense de Telecomunicaciones y Correos (TELCOR). However, TELCOR's powers are limited to demanding information and adopting pertinent corrective measures, which are not specified (Ansorena, 2007). The agency is headed by a Director General appointed by the President of the Republic.

The weakness of the regulatory body was accentuated by the serious conflict between the President of the Republic (2001–2006) and the alliance of Sandinistas and Liberals.21 The latter used their parliamentary majority to introduce important reforms that affected the Supreme Electoral Council (CSE), the CSJ and the Office of the General Comptroller of the Republic. In this context, and with the opposition of the President of the Republic, the above-mentioned political alliance succeeded in passing the Law creating the Superintendence of Public Services (SISEP) on November 23, 2004, in the Legislative Assembly. This new entity had to bring together in one single government structure the regulatory bodies of the sectors providing basic services (water, electric power and telecommunications). This approach to the institutional regulatory framework could be positively assessed in the context of the international discussion, bearing in mind the shortage of human resources in the country. However, the fact that at least during 2006 both TELCOR, supported by the President of the Republic, and SISEP, supported by the National Assembly, were operating gave rise to an unsustainable situation that made promotion of competition in the telecommunications sector even more difficult.

The above analysis makes it clear that the countries of the region that have no competition authority suffer from an institutional weakness in their regulatory framework, which, in turn, is an obstacle to create favourable conditions for competition.

What has happened with Mexico and Panama, which faced the opening up of the industry to competition when they already had a competition authority?

From the point of view of the sequence, Mexico operated in a very different manner from that usually recommended. It privatized in 1990, organized the Federal Commission on Competition (CFC) in 1994, passed the LGT in 1995, created the Federal Telecommunications Commission (COFETEL) by means of a presidential decree in August 1996 and opened the market to competition in 1997.

If analysed in general terms, the LGT proves adequate. It defines a regulatory framework and includes standards that assume a competitive structure in the industry. Nonetheless, if analysed in greater detail important difficulties appear. First of all, the sequence followed by the process was wrong. The concession licence granted to TELMEX in 1990, which regulated the performance of the enterprise until the approval of the general law, was likely to collude with the latter and with the Competition Law. Indeed, the LGT provides that licencees could freely set the rates for telecommunications services, which should be non-discriminatory. Only if a company has substantial power in the relevant market according to the Federal Law on Economic Competition (LFCE) can the Ministry of Communications and

21 Enrique Bolaños formed part of the Liberal Party headed by ex-president Arnoldo Alemán (1996–2001). However, both representatives drifted apart when Bolaños's government joined the accusations of corruption against Alemán (who was imprisoned for that reason).

Transportation (SCT) establish specific obligations related to rates, service quality and information (Article 63). In 1997, the CFC declared that TELMEX had substantial power in five markets (local telephony, access, inter-urban transport, national LD and international LD) and consequently requested from COFETEL the establishment of specific obligations regarding rates. TELMEX complained and held that such statement violated its concession licence.

A second difficulty was that the LGT assigns the SCT the basic regulatory authority. In spite of this, in August 1996, President Ernesto Zedillo created COFETEL by presidential decree, which basically plays an advisory role to the SCT, to which the law assigns the regulatory functions. Thus, a highly ambiguous situation arises. Formally, the regulatory body is COFETEL, but due to its weak legal base, companies question the legitimacy of any decision adverse to them. In fact, on crucial issues the SCT, particularly the Under Ministry of Telecommunications, is the regulatory body. Like many Executive Branch agencies, the under ministry tends to privilege short-term growth of investment and coverage, even at the cost of the development of competition (Mariscal and Rivera, 2007).

A third difficulty was that even though the Telecommunications Law includes among its objectives fostering healthy competition among the different providers of this service, no institution is assigned this generic function. Indeed, among the functions of the SCT, no specific mention appears in this regard; nor does the Federal Law on Competition appear among the laws that should be applied if no express provision on competition appears in the Telecommunications Law. Only in the case of tenders does it indicate that the bases should include a favourable opinion from the CFC. It is the decree creating COFETEL that said agency is obliged to register the rates for telecommunications services and establish specific obligations with regard to rates, service quality and information for licences of public telecommunications networks that have substantial power in the relevant market. Naturally, an Executive Branch decree is a weak legal instrument to assign powers of such importance to an administrative agency.

The above difficulties are compounded by the non-existence of a fixed term of office for COFETEL commissioners and of the specific causes of removal from their posts. Particularly important is the generalized use of the remedy of amparo which allows the regulatory action to be suspended when those affected consider that the norm applied violates their constitutional rights. In a context in which justice operates slowly, these remedies can remain in process for as long as three years, which naturally leads to ineffective regulation (Mariscal and Rivera, 2007).

In Panama's case, even though nominations for commissioners take place in the manner indicated by the Law, it seems it has not been altogether possible to prevent them being carried out on the basis of short-term political interests. In fact, the predominance of the party in government can be accompanied also by its prevalence in the Legislative Branch and, in this context, make politically based nominations, not in line with the capability required by the post (González, 2007). A similar problem has been put forth in the case of Nicaragua (see Ansorena, 2007).

In Panama, the fact that the state retained 49% of INTEL's shares gives rise to a conflict of interests since the government expects high earnings for the operator who can provide higher public revenues, but this can hinder the regulator's mission of applying an appropriate tariff policy (González, 2007).

III. From Exclusivity to Competition: Models and Main Lessons

As we have seen throughout this chapter, the countries of the region that privatized and opened up the telecommunications industry to competition show dissimilar processes. The sphere of design of the institutional economic framework also differs among countries.

However, all the countries studied, whether or not they went through a period of exclusivity, have opened their doors to a competitive model. The aim of this section is, first, to analyse the principal characteristics of the opening up to competition, identify the positive aspects and problems of each experience and subsequently, in the fourth section, evaluate the industry's performance in each of the countries and telecommunications segments.

As we saw above, three countries, Nicaragua, Mexico and Panama opted for granting a period of exclusivity to the new owners of the privatized enterprises. These periods varied in length (six years in Mexico and three years in Nicaragua and Panama) but in all of them, exclusivity included fixed and LD telephony and gave the company that acquired the public enterprise the right to operate a mobile telephony licence.

Another two countries, Guatemala and El Salvador, opted for immediately opening up the industry to competition after the privatization without offering exclusivity. Before carrying out this operation, El Salvador separated fixed telephony from mobile telephony and sold them to different operators, whereas Guatemala transferred a single enterprise to the private sector in an integrated manner.

1. Exclusivity and Opening Up to Competition: The Cases of Mexico, Nicaragua and Panama

The literature has confirmed that privatization with exclusivity represents significant costs for society as regards both coverage and service prices. Scott Waltsten (2000) empirically demonstrates that a monopoly is more valuable to its owners than a company operating in a competitive market and concludes that a government can practically double the revenues obtained from privatization by guaranteeing exclusivity rights. Coinciding with some of the findings in this chapter, Waltsten's study also shows that the greater revenues are accompanied by a reduction in the growth of the telecommunications network of up to 40%.22

22 A similar conclusion is reached by Estache et al. (2002, p. 157) who, on evaluating the progress of telecommunications in Latin America, maintain that "high prices continue to be perceived as

In Mexico's case, the opening up to competition shows distinct characteristics in each of the segments. In 1987, when the mobile market took off, the country was divided into nine regions, with two licences granted in each one. TELMEX, through its subsidiary Telcel, received a licence in each of the regions, the only proviso being that the company could not be the sole provider in any of those regions. In TELMEX's concession licence, the SCT reserved the right to grant another, or other concessions, in favour of third parties in order to develop, in equal circumstances, fixed telephony and national and international LD services, within the same geographical area or in a different one, identical or similar to those that were the subject matter of the concession. However, during the six years following the date in which the concession licence was signed, the SCT pledged to grant other concessions for national and international LD basic telephony public service networks, only if the company had not fulfilled the conditions of expansion and efficiency set forth in the concession licence, which contradicts the above. Chapter V of the said licence defined the general conditions for interconnection and the company's interconnection obligations with other LD public networks when the sector was opened to competition as of January 1, 1997 (Mariscal and Rivera, 2007).

Between 1988 and 1990, Telcel – TELMEX's mobile affiliate – expanded its cellular network in the 800 MHz (Band B) radio frequency, covering the cities of Tijuana, Cuernavaca, Toluca, Guadalajara, Monterrey and the Mexico City metropolitan area. In 1990, the company began to offer cellular telephony in the nine regions into which mobile telephony service was divided in Mexico. In 1998, the 1,900MHz (Band D) radio frequency for PCS in the nine regions was awarded to Telcel. Subsequently, in 2004, a wide-ranging auction was carried out that would make it possible to meet the country's enormous demand for mobile telephony. Towards the end of the last decade, the segment's main operators, together with Telcel, were Iusacell, operating since the end of 1980s, Pegaso which entered in 1999 and Unefon which began operations at the beginning of the year 2000. The entry of Movistar, of Telefónica España, in 2002, contributed dynamism to the market. Nevertheless, up to 2004 the share of companies other than Telcel reached only 22.8% (Mariscal and Rivera, 2007).

By 1994 applications to enter local telephony had already been presented. Iusacell, a joint venture between Alejo Peralta and Bell Atlantic, proposed providing the service by means of fixed wireless technology, but it was not until October 1997 that the rules for the development of competition in local telephony were published. The objectives of the said standards were to promote competition,

an issue in the region. To a large extent this is a result of the limited competition in the sector. For most countries, the exclusivity periods granted to get the privatization deals done, resulted in lasting high connection and usage tariffs. ... Indeed, residential connection rates continue to be high when compared to the US even if they have gone down significantly. These exclusivity periods are now coming to an end as in Argentina or Venezuela and this should lead to a reduction in tariffs driven by the market. However, most countries have not yet defined the rules of the game to ensure competition in a sector where costing the access to bottleneck facilities continues to be at the core of the regulatory debate."

facilitate interconnection and interoperability of public networks, ensure the continuity of the service and avoid the application of discriminatory rates. In 2003, eight companies were authorized to provide local wired service (Ramírez, 2005). The results, however, are not encouraging. In 2004, seven years after the segment was opened up to competition, TELMEX maintained a share equivalent to 95% in local telephony, although the companies that entered the market did increase their share in a sustained manner (Mariscal and Rivera, 2007).

In June 1994, the SCT issued the "Resolution on the Plan for Interconnection with Public Long-Distance Networks" which, among other things, obliged TELMEX to interconnect its networks with all the interested parties. It was expected that by early 1997, 60 interconnection points would be in operation and that they would increase to 200 in the year 2000. With the prospects of the opening up of LD to competition, several companies began to apply for concession licences to operate in this segment.23 In January 1996, the first negotiations among the operators took place, without agreement being reached on interconnection rates, which led them to request the intervention of the SCT. Both TELMEX and the new LD operators questioned the SCT's resolution by means of formal claims. Whereas for TELMEX the charges were insufficient to cover the costs it had incurred to provide the service, for the entrants there were no reasons to justify them. Following a year of analysis, the recently established COFETEL upheld the above-mentioned rate, although according to different international standards it was extraordinarily high.24

In January 1997, seven companies entered the LD market, users being able to pre-select the operators in 60 cities. Among the licencees were Mexican industrial and financial groups associated with US telecommunications operators. Three major operators dominated the market: TELMEX, which established a joint venture with US Sprint to provide international services between Mexico and the USA, Avantel (Banamex and MCI) and Alestra (Alfa and AT&T), which merged with Unicom (Bancomer, GTE and Telefónica). Despite the problems mentioned above with the opening of the market, the companies that entered quickly gained shares in national LD, attaining 30% in 1999. Since 2001, this process has gradually reverted, dropping to 23.2% in 2004.

Seventeen years after opening up to competition in mobile telephony and almost nine years in local and LD telephony, the results from the point of view of operators other than the incumbent were not encouraging. Compared to other Latin American

23 Together with TELMEX, which had a licence since December 1990, Avantel obtained it on October 6, 1995; IUSATEL on February 12, 1996; MARCATEL on February 26, 1996; INVESTCOM on April 4; BESTEL on April 10; MIDITEL on April 26; ALESTRA on April 30 and TELINOR on July 23.

24 A rate was established for cities with Interurban Traffic Centre (CTI; US$0.0257 in 1997 and US$0.0231 in 1998) and another for cities with Routing Capacity Centers but without CTI (US$0.0240 in 1997 and US$0.0219 in 1998). Furthermore, the SCT determined that the companies should pay an additional 58% for traffic termination, which raised the effective rate in 1997 to US$0.0536 and US$0.0513, respectively. By contrast, a World Bank study of the Mexican market suggested an average weighted rate of US$0.019 cents.

countries, Grupo Carso's share of the three segments analysed in Mexico far exceeded the incumbents' share in Argentina, Brazil, Chile and even Peru. It should be pointed out that this situation has occurred despite the fact that it was the companies from the USA, Mexico's great neighbour, which tried to become consolidated in the Mexican market, but without success.

One important specific element has to do with the modalities provided for in the Law for the establishment of interconnection and access charges. As in other countries, the LGT assigns the operators themselves the task of agreeing on those charges. What can be a reasonable, efficient mechanism in a context in which the companies are of a relatively equivalent size is not when companies not only of different sizes but also having strictly contradictory interests come together in the negotiation process. Indeed, as soon as the exclusivity period is over, the entrants face the urgency to begin operations. The incumbent, on the other hand, is interested in hindering the entry of its competitors as much as possible, and, furthermore, every day that the negotiation process is lengthened is another day that it continues to have exclusivity. The more the negotiation period is prolonged, the longer the time during which the entrant has to face its fixed costs and its financial costs without beginning to receive income, which clearly stands as a barrier to the entry of competitors. The delay in the start of operations significantly raises costs for competitors, reducing their competitive possibilities.

Representatives linked to TELMEX and to COFETEL maintain that the entrants' difficulties stemmed from their low levels of investment, from the limited knowledge of US companies and other foreign operators of the characteristics of telecommunications markets in developing countries and from errors on the part of the executive bodies (Mariscal and Rivera, 2007). Nonetheless, the characteristics of the privatization process with a long period of exclusivity that consolidated a single operator and the weakness of the institutional regulatory framework certainly played a central role. It can be said that the institutional framework did not work properly. The various complaints filed by the different competitors both before the CFC and COFETEL took a long time to be resolved, losing all effectiveness. The declaration of dominant enterprise that the LFT provided for as an instrument to avoid anti-competitive conduct by the incumbent company had no effect due to the said way in which the declaration was written and the remedy of amparo which neutralized any administrative measure by the regulatory body and the competition supervisor. Everything points to problems with the legal framework which go far beyond the sphere of regulatory policy and protection of competition (Mariscal and Rivera, 2005a, b).

In the Panamanian model of transition to competition, the opening up was due to be implemented in two phases: the first, corresponding to basic telephony, on January 2, 2003; and mobile telephony in January 2008. The regulatory framework included the following model for opening: the point of departure was acknowledgement that the entrants would face serious difficulties resulting from the consolidation of the incumbent enterprise. In this regard, the new companies were assured interconnection provided by the incumbent in an (apparently) peremptory period and total freedom to choose the services to be provided, as well as the customers. It was believed that by choosing the most profitable markets, entry would be made feasible. The unbundling of networks was also provided for, i.e. the possibility for entrants of leasing segments of the network from the incumbent to provide their services and number portability (González, 2007).

The opening was prepared well in advance. As of early 2002, numerous companies applied for concessions. Up to February 2005, there were 28 concessions for local telephony, 33 for national LD, 59 for international LD, 20 for public telephones and 13 for the hiring of dedicated voice circuits. The basic idea was for the entering operators to have their installations ready to enter into operation on the appointed date. Bellsouth, a company that controlled 45% of subscribers, requested concessions for all the services. Cable Onda, S.A., the dominant cable television operator, did much the same. It was therefore a question of operators with a significant market share and a solid foundation in the new telecommunications technologies.

Despite the existence of a regulatory framework favourable to the entrants, the liberalization of the sector has encountered obstacles that in the best of cases have substantially delayed implementation. There are four fundamental aspects. (i) The interconnection mechanism, in which the regulator only intervened if the parties could not reach agreement, contained incentives for both parties to await the regulator's intervention, which substantially delayed the processes. Even having defined the conditions and the interconnection rate, the problems did not end, for numerous technical difficulties have further delayed the process.25 (ii) With regard to competition in LD, problems arose both in the use of access codes and in automatic routing. The changes demanded by the use of various access codes by users were subject to the speed at which the incumbent implemented them. Regarding the second mechanism, it was assumed that its operation gave rise to change costs, basically derived from the programming the incumbent had to assume. If these costs were to be covered by the user, the change of provider would be discouraged. Alternatively, if these costs were to be assumed by the entrant, its competitiveness would be reduced. (iii) Related to the above, the survey had to be carried out among users so that they choose their LD services provider, even though done as scheduled, mainly because the cost of the calls could not be determined before the negotiations on interconnection charges had concluded. (iv) Finally, number portability had still not been put into operation and judicial proceedings halted efforts regarding unbundling of the network. With regard to the latter aspect, the non-existence of an explicit standard in this respect, within the legal framework, led this policy to be questioned before the courts (González, 2007).

25 The most significant problems that arose were the expansion of routes and new interconnection points. In the case of the expansion of the E1 for POI in the city of Colón and in the provinces, as requested by the entrants, this had still not been achieved until June 2004. Thus, to that date, competition in national LD had not yet begun, despite the fact that, as we saw, there were a significant number of entrants according to the regulatory framework; it had begun a year and a half earlier. For a detailed description of the perverse incentives of the regulatory framework and the "technical competitive" problems encountered, see González (2007).

Panamanian legislation does not encourage the incumbent to allow interconnection, as it would threaten its monopolistic position; nor does the entrant have incentives to reach agreement with the incumbent, since, having analysed the incumbent's negotiating capacity, it is unlikely that the latter would be prepared to grant a lower access charge than the regulatory body would define, and it is therefore in its interest to await the regulator's arbitration. Finally, the regulator has strong incentives to "split the child in half" since according to Article 216, interconnection charges should reflect "at least the long-term incremental costs" (González, 2007). Having added the practical problems,26 the consequence of the above was that in mid-2004, i.e. a year and a half after the opening up, important cities such as Agua Dulce, David, Chorrera, Santiago and Chitré did not have interconnection points.

Despite the above, opening up to competition in LD has meant a substantial drop in prices due to both the impact of the threat of competition and the lessening of difficulties in implementing the required investments. According to estimates the average minute of national LD was to drop from $0.15 in 2002 to $0.07 in 2007. In the case of local telephony the progress has been negligible.

This situation raises a more general debate. Whereas some placed emphasis on the unbundling of networks as an important mechanism for introducing competition in fixed telephony, others maintain that it is an unworkable policy (the owner of the network cannot be compelled to lend their network and even less below its cost) which discourages investment (the entrant, if it can use the incumbent's network, will not want to invest). The Panamanian experience appears to show three basic problems: (i) the difficulties in moving from a system of exclusivity to one of competition are usually underestimated; (ii) the capabilities of network unbundling as a basic instrument for introducing competition in fixed telephony are overestimated; and (iii) the postponement of full liberalization of mobile telephony until 2008 hinders the most effective means of introducing competition throughout the sector – mobile telephony.

As we saw, the third country that granted a period of exclusivity to the former public enterprise when it was privatized was Nicaragua. The privatization process culminated in December, 2001, and therefore the opening up to competition was to begin towards the end of 2004. However, the process was postponed until April 2005, by means of a parliamentary resolution.27 Subsequently, the conflict between the two regulatory bodies, TELCOR and SISEP, had translated into hard bargaining regarding whether the telecommunications should be opening up or not. Whereas

26 The most important problems were those of expansion of routes and new interconnection points. In the case of the expansion of the E1 for the POI in the city of Colón, this had not been achieved in June, which hindered competition in national LD. The ensuing interconnection problems translated into problems of congestion, which were a serious disincentive for consumers to change provider (González, 2007).

27 The National Assembly established that the opening did not have to be on December 18, 2004, three years after the signing of the privatization contract, but had to be three years after ENITEL's concession contract was published in the Gazette, which was in April 2005.

the first entity declared the total opening up of this sector in April 2005, the recently created SISEP disregarded such measure and proposed an extension of ENITEL's private monopoly and a freeze on telephone rates after the start-up of the new regulatory body (Ansorena, 2007). This situation resulted in serious disorder. Towards the end of 2005, ENITEL's monopoly in basic telephony continued, SISEP was engaged in preparing a new project for opening in consultation with the operators and the mobile telephony company had entered the international LD telephony segment, refusing to recognize ENITEL's exclusivity (Ansorena, 2007).

As to mobile telephony, Bellsouth operated in conditions of exclusivity between 1997 and the end of 2002. In 2001, TELCOR decided to open the segment to the participation of other operators by initiating the bidding process for the second mobile telephony licence. Thirteen companies took part in the bidding. The operating licence was finally28 awarded to PCS SERCOM, an affiliate of the Mexican company América Móvil, which began operations in December, 2002, simultaneously with ENITEL, which had the right to a mobile telephony licence in accordance with the privatization contract. Competition in the segment suffered a grave setback when the above-mentioned Mexican company acquired ENITEL in 2004, and thus held two mobile telephony licences. The non-existence of a competition agency that could have established as a condition for the purchase of ENITEL that América Móvil should get rid of one of the mobile licences represents a serious problem for the development of competition in the sector. Bearing in mind that mobile companies are central to the development of overall competition in telecommunications,29 the prospects for competition in Nicaragua are not at all promising. Despite the foregoing, it should be pointed out that the number of mobile subscribers rose from 200,000 in 2002 at the end of the exclusivity period to 686,000 in 2004.

2. Privatization and Opening Up to Competition: The Cases of El Salvador and Guatemala

In contrast with the countries analysed in the preceding section, both El Salvador and Guatemala rejected the possibility of granting a period of exclusivity to the companies that acquired the former public enterprise and opted instead for immediately opening up the industry to competition.

In El Salvador's case, the decision to sell the assets of the fixed company separately from those of the mobile company is also noteworthy. Thus, the presence of at least two telecommunications operators was assured. The option for a

28 In fact, the bidding was won by another operator linked to president Alemán, the owner of which was later accused of corruption. The licence was transferred to the second bidder, PCS SERCOM.

29 On the importance of mobile companies in overall competition in telecommunications, see Mariscal and Rivera (2005a, b).

competitive market in mobile telephony has translated into the fact that despite the country's small size, there are four companies operating: CTE-Telecom (since 1999), Telefónica (since 1998), Telemóvil (since 1992) and Digicel (since 2002). Three companies have national coverage and one of them has centred on the urban area (Digicel).30 After having only one operator in 1998, by 2004 the mobile market was distributed fairly equitably among the four operators.31

As Argumedo (2007) points out, there is another situation in fixed telephony, where CTE maintains its dominant position. In 1998, the company had 386,600 lines; by 2002, the total number of lines had risen to 709,400, of which CTE had 90%; and by the end of 2004, the lines had continued to grow, although at a slower rate, reaching 887,800. CTE's share continued to be high and was equivalent to 88.7%. According to SIGET, there are another nine companies in fixed telephony, but the concentration is greater than in mobile telephony. CTE continues to be practically the only company that offers fixed telephony services to the residential segment, and it faces more competition in the segment of service to businesses.

In Guatemala the process of interconnection between the incumbent operator and the new operators began in mid-1999, and conflicts arose from the start. In the mobile segment, when the "caller pays" mechanism was introduced in October 1999, there were differences of opinion and legal disputes over this system. There were three cases of arbitration for interconnection charges in mobile networks. In international LD there were also tensions due to the fact that, among other things, the earnings from incoming international traffic included subsidies to local telephony. There have been many conflicts over interconnection charges for incoming international traffic, two of which went to arbitration, but those regarding international termination charges continue to be in conflict. The need to gradually eliminate this situation in the context of the interconnection contracts signed has also given rise to other disputes between operators, caused by: (i) blockage of calls to operators who have filed complaints before the SIT; (ii) lack of renewal of interconnection contracts with the incumbent, which has led to three complaints before the SIT; (iii) discriminatory rates for calls to competitors, giving rise to complaints in ordinary proceedings; and (iv) double charges to users of international LD, among others (Urízar, 2007).

The LGT does not establish any restrictions on the entry or exit of competitors, but most operators,32 some international analysts33 and SIT officials, believe that in practice the barriers to entry come from discriminatory treatment and the difficulty

30 According to Argumedo (2007) the competition is intense. The companies offer varied plans to attract different segments of the contract market (individual, family, friends, corporate) and prepaid cards for different amounts. They also offer a wide variety of complementary services such as roaming, text messages, sending of photos and sending of emails. The companies have been very aggressive in expanding the network of points of sale and charges for products and services.

31 Even though Telemóvil keeps a greater participation (28%), CTE, Digicel and Telefónica, follow closely, with 28%, 22% and 21%, respectively (Argumedo, 2007).

32 Interview with Mr. Baldir Garrido, President of the Union of Telecommunications Operators, March, 2005 (Urízar, 2007).

33 Ibid (p. 11).

that the incumbent imposes on other operators to access and interconnect to public networks. This difficulty is generally reflected in: (i) delays in interconnection negotiations; (ii) rates for termination of international calls traffic being higher than those of the market; (iii) difficulty in making payments in cascade between the different service providers, since the latter have agreed that to make telecommunications transfers between companies that do not have direct interconnection, the operator who carries out the function of traffic only will transfer the traffic if there is an interconnection contract between them. (iv) Another reason many operators experience difficulty in interconnection and access to essential resources is that the list set forth in Article 27 does not include key elements for the development of effective competition in the sector, such as the unbundling of infrastructure, joint location of equipment, rights of way and sole invoice service, among others. Thus, for example, operators have come up against difficulties in obtaining rights for the placement of posts or ducts, which are governed by the regulations of the different municipalities and are therefore subject to the discretionary authority of the mayor. (v) Another aspect that has hampered interconnection is that the incumbent operator offers only two interconnection points in the municipality of Guatemala, making entry difficult, since the entering operator must pay for transport of telecommunications to those points, even if it possesses the infrastructure to transport the communications by its own means (Urízar, 2007).

The de facto opening up to competition in Honduras promoted by the TpT has not responded to expectations, despite the progress made. In fact, when the project was launched, it was expected that towards the end of 2005, 200,000 lines would have been installed. By September only one third of these had been installed. Among the main reasons for this situation were: (i) initial interconnection problems that obliged sub-operators to wait approximately seven months after the programme was launched before the first of them went out on the market; (ii) lack of definition and selection of the technology to be used by sub-operators, and many of those that obtained registration as sub-operators acted more as a result of the momentum generated by the project or of false expectations of the telecommunications business than a thorough knowledge of the business (Tábora, 2007). An interesting aspect of the Honduran experience has been the technological variety applied by the different companies.34

IV. Industrial Organization and Competition in Telecommunications in Central America and Mexico

In the preceding section the paths followed by the opening up to competition in the different countries were compared. It was shown how the countries that accompanied privatization with the granting of a period of exclusivity have had more difficulties than the other countries in creating a competitive market. However, the non-existence

34 At the beginning of the programme, most of the interested parties focused on the use of wireless systems; however, due to the limited spectrum availability many rejected this system. It is worth

of institutions for the defence of competition and the inexperience and limitations of regulatory bodies have made the transition to competition difficult even in the countries that did not grant a period of exclusivity. The purpose of this section is to evaluate the forms acquired by the industrial organization of the sector and compare the different models and their results in terms of number of operators per segment and the dynamism of the different markets.

As in the rest of Latin America, the telecommunications market in Central America has become the object of dispute between the two dominant operators in the region, Telefónica España and the Mexican group TELMEX – América Móvil, part of the huge Grupo Carso.

Since the early 1990s, Telefónica España has been developing an aggressive policy of expansion in the telecommunications market in Latin America. This policy, which began in South America, has been progressively expanded to the north of the continent.35

In fact, Telefónica began mobile operations in El Salvador and Guatemala. Despite its interest in participating in the bid for INTEL of Panama, it was unable to do so, since that country's legislation prohibited the participation of companies in which foreign governments had a majority share.36 Subsequently, after Telefónica reached an agreement with Bellsouth in March 2004, to acquire all of its operations in Latin America, Telefónica took control of Bellsouth's operations in Guatemala, Nicaragua and Panama. In parallel, the company began operating in Mexico in 2002 by means of various acquisitions until it became the second operator in that country after Telcel. Thus, the company began to have an increasing presence in the territory of its major competitor at regional level: the TELMEX – América Móvil group.

This latter group, which originated in the privatization of TELMEX in 1990, had concentrated almost exclusively on its operations in Mexico until 1997. The backward state of telecommunications in that country, its enormous size and the threat posed by US companies, particularly in the context of NAFTA, justified the exclusive attention that the company gave to its country of origin. This policy began to change drastically at end of the 1990s.

1. Competition in Fixed Telephony

The problems of competition in this segment are evidenced by the incumbent's high share in these countries. Those which provided an exclusivity period have experienced greater difficulties in moving towards competition.

highlighting the variety of technologies used by sub-operators to undertake their projects: satellite and wireless access such as Teléfonos de Honduras; wireless access such as MULTIFON, UNITEL, CELTEL and Community Telephony, among others; copper networks such as INTELDATA; cable subscription television networks such as SULATEL, AMNET and Cable Color, among others; data transmission and switching networks such as METRORED; Informatics Access Networks (Internet) such as ITTSA, to mention a few cases (Tábora, 2007).

35 For a detailed analysis of this process, see Mariscal and Rivera (2005a, b).

36 As is known, Telefónica from Spain was completely privatized.

Up to 2004, TELMEX of Mexico had a 95% share of fixed telephony. Unfortunately, cable television operators had not developed the so-called Triple Play and hence, there has been practically no competition in broadband.

In Panama's case, entry to the fixed telephony segment has been very difficult. Even though by February 2005, 28 concessions for local telephony had been turned over, competition has concentrated on national and international LD and the business segment of local telephony. The difficulties encountered by the interconnection process have delayed the effective entry of competing companies, and therefore their presence is still minimal.

In Nicaragua, the control package was acquired by the consortium Telia Swedtel Ab – MEGATEL EMCE in 2001, while the state pledged to sell the remaining 49% it owned in a period of three years. In 2004, the privatization process was finally completed with the sale of the shares to the Mexican company América Móvil. That same year, seeking to expand its position in the regional telecommunications market, América Móvil took over the operator MEGATEL of Honduras, thus becoming the only shareholder of the Nicaraguan operator ENITEL. As mentioned above, the exclusivity period was expected to finish towards the end of 2004, despite a number of problems, among them the "competition" between the two telecommunications regulators, which was a sign of the country's serious political conflict. Yet, in 2006, ENITEL continued to be the monopolistic provider in fixed telephony (Ansorena, 2007).

In the case of Honduras, the end of the exclusivity period was scheduled for the end of 2005. The TpT programme, with the introduction of the concept of sub-operator, made it possible to move this process forward. By August 2005, according to CONATEL data, 40 companies had been registered as sub-operators. Of these 40 companies that complied with the requirements established in the process, 17 have begun commercial operations and it is estimated that to date they are operating approximately37 63,000 telephone lines commercially, around one third of which belong to only one sub-operator.38 Taking this information as a reference, to December, 2003, 100% of the fixed telephony market was in the hands of HONDUTEL. By December, 2004, with the entry of sub-operators this percentage dropped to 93.69%, the remaining 6.31% being in the hands of sub-operators. In August 2005, sub-operators had 13.83% of the total telephony market, whereas HONDUTEL maintained dominance with 86.17%. Three aspects should be underscored in this context. First, the great success achieved by the entrants, which contrasts with the cases mentioned before. Despite this progress, 98% of the lines installed were located in the country's three most important cities, deepening the coverage inequity between urban areas and rural areas, thus leaving the government to attend the social problems of access to telecommunications services (Tábora, 2007).

37 Even though the regulations issued by CONATEL for the programme "Telephony for All – Modernity for Honduras" establish compulsoriness in delivery of information on each operator's new subscribers, no evidence was found that this requirement was being complied with, and therefore the data on subscribers are estimated, based on the Tabora's own research.

38 MULTIFON, S.A. de C.V.

This situation confirms the incumbents' repeated claims about the policy of "cream skimming" developed by the entrants. Finally, the presence of a public operator who shares the government's interest in promoting the entry of other operators may have been a contributing factor in reducing conflicts with regard to interconnections. Furthermore, it should be taken into account that at least until 2006, sub-operators had not had to pay charges for the use of HONDUTEL's network. This regime was transitory and concluded on December 25, 2005. As of December 26, 2005, the regime of payment of access charges established in the Regulations for Interconnection was enforced (Tábora, 2007).

What happens in the case of countries which, together with privatizing the enterprise, opened up fixed telephony to competition?

In El Salvador ten companies participated, among which the most important were CTE-Telecom, Telefónica, Emetel, GCA Telecom, El Salvador Telecom, El Salvador Network, Telemóvil and Newcom. However, as we saw in the previous section, CTE's share continued being large (88.7% of the market). In spite of this, the immediate opening up to competition has allowed greater participation by entrants.

An analysis of the different national cases shows that over and above the differences displayed by the privatization processes, the transition to competition in fixed telephony has proved more complicated than expected. Indeed, beyond the intentions what the processes in the region show is that a public monopoly has become a private one.

If we analyse the fixed telephony industry from a regional point of view, the situation is even more complicated. As we can see in Table 3.1, in four of the five countries that have privatized the public telecommunications enterprise, the TELMEX – América Móvil group appears as the owner of the dominant enterprise in fixed telephony. Only in Panama is the dominant company owned by another group. The above-mentioned table also makes it clear that there is no other fixed operator with any capacity to generate competition in this segment. In this sense, as the regional telecommunications market consolidates, it is difficult to expect a competitive market to be generated by the countries' dominant companies. On the contrary, only the consolidation of the dominant position of Grupo TELMEX – América

Table 3.1 Central America and Mexico: fixed operators and market shares, 2004 (percentages) (Own preparation)

Móvil would seem possible. We will return to this topic in the following section on analysing the situation of mobile telephony.

In this context, it is likely that if HONDUTEL is privatized, Grupo TELMEX – América Móvil will have very strong incentives to seek to win the bid. Similarly, should telecommunications be opened up in Costa Rica, the possibility of ICE becoming a regional operator, taking advantage of its capabilities and the synergies created by the fact that it is also an electric power enterprise, is not very feasible.

Table 3.2 shows the evolution of the rate of penetration of fixed telephones in the seven countries included in our study. The excellent performance of fixed telephony in Costa Rica, the only country that has flatly rejected privatization of the enterprise, is one of the outstanding phenomena. Compared to Mexico, in 1990 Costa Rica had a penetration 1.55 times that of Mexico; in 2004, the difference widened to 1.84 times. A second interesting aspect is that the growth of coverage in fixed telephony in Costa Rica is very dynamic, between 1998 and 2004 it rose from 19.33 to 31.62.39

Mexico, in contrast with Costa Rica, shows an historical low penetration of fixed telephony (6.48 in 1990) and a very slow expansion of its coverage. It was only in 1998 that it reached 10.36. Interestingly, it was as of that year, one year after the opening up to competition, that the growth of teledensity accelerated. As in Costa Rica's case, although at lower levels, the growth of fixed telephony continues coming close to the South American countries, whose teledensities have a tendency towards stagnation. Nevertheless, compared to the level reached by per capita GDP, Mexico's penetration rate is low (see Fig. 3.1).

Table 3.2 Mexico and Central America: fixed telephone lines per 100 inhabitants, 1990–2004 (International Telecommunication Union (ITU) Online)

39 As is known, there is a generalized tendency in Latin America, despite the low teledensity, for fixed telephony to stagnate, and even to deteriorate in recent years.

Fig. 3.1 Central America and Mexico: penetration in fixed telephony (International Telecommunication Union, 2005)

The cases of El Salvador and Panama are particularly interesting due to the contrasts shown. Whereas in 1990 teledensity in El Salvador barely reached 2.42, in Panama it stood at 9.27, just behind Costa Rica. In 1997, when Panama's public enterprise was privatized, the penetration rate was 13.44; in 1999, this rate reached a maximum level of 16.43; and the following year it began a continuous decline until in 2004 it reached a teledensity of only 11.85. In stark contrast to this, in El Salvador teledensity grew continuously until it surpassed Panama in 2004, when it reached a figure of 13.42. These different paths cannot be explained by uneven growth in mobile telephony. In fact, as we shall see further on, the number of subscribers per 100 inhabitants was, to 2004, fairly similar in both countries, although El Salvador slightly surpasses Panama. The differences suggest, rather, that the competitive atmosphere favourably affects rates and this has a positive impact on teledensity.40 The performance of the other three countries has been very modest. In 2004, the level of penetration in Guatemala was marginally over 8%. Honduras was barely approaching 6 points. In Nicaragua, the benefits of privatization had simply not become apparent.

Table 3.3 shows a synthesis of what has been discussed in this section and helps to evaluate the development of fixed telephony over the last 15 years. The countries which have liberalized the telecommunications sector, after an exclusiveness period, are characterized by a greater participation of the incumbent than the one it has in those countries which simultaneously liberalized and privatized this sector. Mexico is, in this sense, a paradigmatic case because the main operator had 95% participation

40 The next section shows that whereas rates stood at an average of US$0.06 for 3 min in El Salvador, in Panama they reached 12 cents.

Table 3.3 Central America and Mexico: privatization, liberalization and performance trajectories, 1990–2004 (percentages) (Author, on the basis of official figures)

even eight years after the market opening up. In contrast, in El Salvador and Guatemala the participation of the main operator reaches around 15%.

If the coverage is considered, Mexico appears with a low one, at least in comparison with Costa Rica. In fact, its higher income per capita does not show in its teledensity. The setback experienced by Panama was already mentioned. As to prices, these are higher for the 3 min call in those countries which have carried out privatization with an exclusiveness period. Table 3.3, illustrates that the average growth rate during the first three years of privatization are substantially higher in those countries that went through a simultaneous privatization and liberalization process.

The Costa Rican experience is unique: the price of local calls is the lowest of the region, has the highest penetration rate and shows a significant coverage increase rate. Even though greater efficiency is not the main reason for this situation,41 Costa Rica certainly gives a different perspective regarding privatization policies and underlines the need for competition regulatory framework that guarantees a competitive post-privatization market.

2. Competition in Mobile Telephony

As in the case of fixed telephony, mobile telecommunications show different forms of development.

In Mexico, the strong presence of the TELMEX – América Móvil group is also registered in mobile telephony. As we can see in Table 3.4, Telcel's share reached more than 77% in 2004. This telecommunications segment, as we saw above, was conceived

41 We will come back to this point in the following section.

Table 3.4 Central America and Mexico: mobile operators and market shares (Author)

from the start by public policy as a competitive sector. Nevertheless, in contrast to the majority of countries, in which there has been a relatively balanced market distribution, in Mexico the dominant company has tended to strengthen its presence. In 1997, Telcel's share was 63.7%, Iusacell's 23% and other competitors 13%. In the following years, Iusacell began to lose market share. An important factor was not including prepayment among its products. The company was acquired by the consortium made up of Vodafone, one of the world's main mobile companies with operations in more than 25 countries, and Verizon. In 2003, however, these companies abandoned their operation in Mexico after experiencing significant losses. The entry of Telefónica Móviles has revitalized competition in the market. The stagnation in entrants' share continued despite the sharp increase in mobile teledensity in Mexico in recent years.

In Guatemala the influence of the dominant fixed company has also been felt in the industrial structure of the mobile segment. Thus, Grupo TELMEX – América Móvil has attained a share of almost 53% of total subscribers. Nevertheless, the presence of important international companies such as Telefónica and Millicom (see Table 3.4) makes it possible to foresee competitive development towards the future.

El Salvador is undoubtedly the country that shows the promising development of competition in mobile telephony. Indeed, as can be seen in Table 3.4, four operators share the mobile market in a fairly balanced manner.

In Honduras, the original strategy at the time the LMT was approved was to have three operators in the short term: (i) Telefónica Celular, S.A. (CELTEL); (ii) the winner of the capitalization process of HONDUTEL; and (iii) the winner of the bidding for Band B, on 800 MHz (Tábora, 2007). However, the consequences were the difficulties encountered to incorporate a private investor in HONDUTEL and the failure of the bidding for Band B until 2002. On January 6, 1994, the concession for development of Cellular Mobile Telephony Services in Honduras was awarded to the companies Motorola, Inc., Millicom International Cellular, S.A. (MIC) and Proempres, S.A., represented by the company CELTEL. This agreement was signed by both parties on August 7, 1995. The concession granted CELTEL the right to develop mobile telephony services within Honduran territory, using frequency Band "A," as of June 1996. The company officially began its services on September 15, 1996. In 2006, CELTEL offered its services in 13 departments and in the country's ten main cities. Until 2003, the number of subscribers grew slowly until it reached 5.5% coverage.

In 2002, CONATEL successfully carried out an international bidding process for Frequency Band "B." Eighteen companies expressed an interest in taking part. In April 2003, three companies were pre-qualified, Empresa Nacional de Telecomunicaciones (ENTEL) of Chile, Bellsouth Honduras and the local consortium, MEGATEL EMCE, which had been jointly awarded with Telia Swedtel the Nicaraguan enterprise ENITEL. However, the first two declined and consequently the company was awarded to the consortium Telia Swedtel – MEGATEL EMCE for a price of $7.1 million.42 The impact of competition on the growth of the number of subscribers has been significant. Mobile teledensity stood at 10.15% in December 2004, reaching a penetration of 14.82% in August 2005 (Tábora, 2007). The performance of competition has been successful. Having begun operations in November 2003, in August 2005, MEGATEL had a share of 31.8% (Tábora, 2007). This may reflect the fact that there are no relations between the incumbent enterprise in fixed telephony and the company that operated the first concession in mobile telephony.

In contrast to fixed telephony, mobile telephony looks more promising from a competition point of view at the regional level. Indeed, three operators have a presence in more than one market, specifically, Telefónica Móviles, América Móvil and Millicom. In addition, other companies such as Cable & Wireless in Panama and Digicel in El Salvador are in operation. Hence, there is an entrepreneurial base that can significantly stimulate competition in this sector.

The presence of several operators largely explains the more dynamic development of this sector. If we look at Table 3.5, it is evident that as of 1998–1999, teledensity began to grow at very high rates. Crucial features of this process are the introduction of the "caller pays" system and prepayment, which made it possible to broaden the services to groups that were not eligible for post-payment programmes. The two countries with the most dynamic growth are undoubtedly El Salvador and Mexico which rose from a penetration rate of 0.68 and 1.82 in 1997 to 27.61 and 36.64 in 2004, respectively.

Costa Rica, in contrast, despite being the country in 1997 with the broadest coverage (except Mexico), in 2004, only surpassed Nicaragua and Honduras. As we shall see further ahead, this appears to be associated with cross-subsidies from mobile to fixed telephony.

V. Telecommunications Rates in Central America and Mexico

For an analysis of prices of telecommunications services in the region a distinction should be made between the fixed segment, subject to more or less strict regulations, and the segments of national and international LD and mobile telephony, under

42 Representatives immediately pointed out that the process should be annulled as only one bidder took part, see Honduras Revista Internacional (http://hondurasri.com/CLON/detalles/TRIBUNA/SER%C1%20ANULADA%20LICITACI%D3N%20DE%20TELEFON%CDA%20M%D3VIL%20PCS.htm).

Table 3.5 Mexico and Central America: subscribers to cellular telephony per 100 inhabitants 1990–2004 (International Telecommunication Union, 2005)

systems both regulated and subject to competition. In these latter cases, one situation or the other is associated with the existence or absence of exclusive concessions.

For a comprehensive analysis of rate regulation for fixed telephony in the region it is essential to analyse the mechanism for initial setting of rates at the time of privatization, the procedure defined for their establishment as a system and finally, the effective path followed by rates and prices throughout the period.

In the case of mobile telephony, its competitive nature, the existence of two separate markets, the prepayment and post-payment markets, the multiplicity of rate plans and diverse variations within each one and constant innovations in promotions make it difficult to follow-up on price policy. Despite the foregoing, this section gives a general overview of prices and rates, seeking to highlight the main characteristics of their evolution in recent years.

1. Rate Regulation and Evolution in Fixed Telephony and Long Distance

In the case of TELMEX in Mexico, the concession licence determines a form of "authorized rate control," i.e. the SCT authorizes the rates set by the companies in line with the following criteria: (i) the rate structure will seek to favour efficient expansion of the public telephone network and lay the foundations for healthy competition in the provision of services; and (ii) the rates applicable to each service should make it possible to recover at least the incremental long-term cost,43 in such a way that cross-subsidies between services are eliminated.44 This is so that there is the necessary incentive to expand each service and establish fair grounds for equitable competition. The rate structure should lead TELMEX to achieve continuous improvement in productivity to enable it to increase its profitability, which in turn should translate into lower rates for users.45

If the tariff rate of growth of a basket of controlled services is compared with the inflation in 1997, the year of the opening up to competition, the former exceeds the latter considerably. Up to 1998, inflation was higher than fixed phone rates, which meant a fall in real rates. This trend continued strongly up to 2002 and 2003, when the nominal rates finally decreased.

43 Average long-term incremental cost means the sum of all the costs TELMEX has to incur to provide a unit of additional capacity of the corresponding service. Incremental costs should be comparable to those of an efficient company, in such a way that regulated rates are internationally competitive.

44 It is understood that there is a cross-subsidy when a company provides a service with an insufficient rate to cover average long-term incremental costs and simultaneously provides another service with a rate higher than its average long-term incremental costs. The above condition will be applied as long as the difference is substantial.

45 For details of the manner in which the concession licence sets forth the price cap system, see Chapter 4.

According to information from COFETEL, the above has translated into significant reductions in the prices of services. Indeed, in June 2004 in the residential sector, the rate for installation had dropped from 1,393 pesos in 1998 to 1,130 pesos in June 2004, representing a reduction equivalent to 18.9%. Basic rent, for its part, had gone from 182.56 pesos to 156.55 pesos, a reduction of 14.2% in the same period. Finally, measured local service dropped from 1.72 pesos to 1.48 pesos, a reduction of 14%. In the business sector, installation costs dropped from 4,991.85 pesos to 3,500 pesos, a 29% reduction; basic rent from 251 pesos to 191 pesos, a 21.1% reduction; and finally, measured local service from 1.72 pesos to 1.48 pesos, a 14% drop, all in the above-mentioned period (Mariscal and Rivera, 2007). In this context, it is noteworthy that the distinction between residential and business service has been maintained, but perhaps what is most important is that the price reductions of these services are not what could have been expected if the more pronounced drops in the cost of equipment and the reduction of unit costs are considered. Additionally the significant increase in the number of users should have helped to further price cuts.

The above appears to be confirmed in Table 3.6, which shows that Mexican local rates exceed those of all the other countries considered in this study. The same occurs when we compare rates in Mexico with those of the other OECD countries. According to a OECD study, up to August 2004, on comparing the costs of a basket of telecommunications services which includes local calls, national and international LD and calls to mobile phones, it is evident that according to the purchasing power parity in US dollars (USD PPP), the cost of the Mexican basket was 611.74, surpassed only by Czech Republic (707.03) and Poland (699.53). The lowest costs were those of Iceland (259.30), Denmark (259.99), Canada (298.29) and Switzerland (306.19).46 The average cost of the basket in the OECD was 428.62 (OECD, 2007).

In El Salvador, maximum rates for public telephone service are determined and approved by SIGET, without detriment to the provisions of Article 108.47 The overall rate for telephone service is subdivided into: (i) access charge, which is the payment

Table 3.6 Mexico and Central America: average cost of local call (dollars per 3 min) (World Bank)

46 It should be pointed out that the small OECD countries have even lower costs for the basket.

47 Article 108 permitted an increase in the rate for fixed telephony based on the operator's investments.

the user makes for being connected to the network; (ii) charge for call traffic (time of use) or services (alarm call, directory enquiries, etc.); and (iii) interconnection charges, plus charges for finishing the call in another network.

Before privatization the rates were set by ANTEL, an institution that between 1980 and 1994 did not adjust the prices charged for traffic in minutes, nor the charges for residential and business access, nor the network installation (FUSADES, 1998), despite the fact that during that period inflation increased by more than 1,000%. The deficits generated were compensated, as was usual at the time, by the income from the high rates charged for international communications. In 1995, the first increase in fixed telephony was registered, increasing the cost per minute of a local call by 200%. The access charge for residential and business telephony rose by 400%, and the installation charge increased by 33.3%. A new rebalancing was applied a year before the privatization, with the rate for a local call experiencing an increase of 200% and the adjustment for the residential charge and installation being lower. International calls made to the USA (the main country of destination) were reduced by 50% in comparison to those of 1994; it should be pointed out that the participation of Americatel was approved in international LD as of the early 1990s (Argumedo, 2007).

The Law established the following mechanism for adjustment of rates: (i) for access charges, a collection base was established which would be adjusted annually in line with inflation; (ii) in the case of interconnection charges, a collection base was determined by type of terminal, which would be adjusted on a quarterly basis according to the inflation rate and devaluation with regard to the USA48; and (iii) in regard to the charge for traffic, a clause was established to allow the rate to be increased as the operator invested more in extending coverage and expanding the network.49 Thus, rate increases were generated despite the fact that they were falling at international level. The continuous increase in rates contributed to the disillusionment of the population with regard to the privatization, since during its planning it had been maintained that competition would make the rates go down (Argumedo, 2007).

If only the inflation rates had been applied, the charge for 1 min access would have reached $0.0185; if, on the other hand, the provisions of the investment contract had been complied with and the maximum increases had been applied, the price would have risen to $0.041/min in 2002 (an increase of 95% over the 1998 price). What actually happened with the rate was that at first the goals for expansion of lines were not met, and therefore the increase authorized was proportional to the effective growth of lines. In June 2002, SIGET and the operator agreed that the increases to which the company had a right were not viable, and they would seek to introduce them in a period of seven years (Argumedo, 2007).

48 In practice the adjustment was only for inflation, since the colon-dollar exchange rate was fixed since 1993 and in 2001 the country implemented dollarization.

49 Based on the existence of 340,000 lines in service in 1997, the investment contract provided for the following rate increases: in 1998 15% on installing 50,000 lines; in 1999 20% on installing 65,000 lines and an annual increase of 20% between 2000 and 2002 as long as 75,000 lines were installed in each of those years.

That is how the rates for fixed telephony have experienced rises, showing the lack of competition and the problems stemming from the investment contract. In the period 1998–2004, there were annual average increases of 4% and 2.5% in basic residential and business charges, respectively. The cost per minute of local calls and national LD, for their part, increased over the past seven years by 5.2% and 4.7%, respectively. The cost of national LD calls reached a figure of $0.04/min, i.e. 40% of the cost of calls to the USA, which could be a sign of market power abuse (Argumedo, 2007).

In the case of Guatemala, the General Telecommunications Law establishes that the rate for consumer, as well as those charged among telecommunications operators, should be freely agreed upon. Government Accord 394–200150 put an end to the actions under Administrative Law Nos. 185–2000 and 186–2000 promoted by the Government against the privatization process. Article 2 establishes that the rates for residential and business services for fixed lines until 2003 should be as follows (using the exchange rate Q. 7.84/dollar): local telephony Q. 0.2036/min ($0.02597/min) and LD telephony Q. 0.3564/min ($0.04546/min).

In 1995, LD rate for calls to the USA, the most important destination for Guatemalan communications, was around $1.50/min, with poor quality and high saturation of calls. In mid-2005, some operators offered rates of $0.10/min.

The rate policy guidelines for Telecommunications Services in Honduras are established in Title Three, Chapter II (Article 31) of the LMT. The latter determines that the rates charged by telecommunications services operators, except broadcasting services, shall be regulated by CONATEL, as long as such services are provided in adequate competition conditions.

Although there was no rate rebalancing in the country,51 the government that took office in 2002 decided, together with CONATEL, to reduce LD rates. The rates effecting place in 2002 for communications to the USA (which represent 91% of all international calls) stood at $1.24/min. That year the rate was reduced to $1.04 and in 2003 it was reduced once again, this time to $0.84.52

50 Diario de Centro América. October 5, 2001–Number 62–Publication of Government Accord 394–2001 of October 1, 2001, Guatemala.

51 Since April 6, 1999, the date on which CONATEL approved HONDUTEL's List of Rates, there has been no rate rebalancing whatsoever. Article 19 of the Regulations on Costs and Rates for Telecommunications Services establishes that there is a period for carrying out the rate rebalancing, which closes on December 31, 2005. The idea of this period was to comply with the contractual conditions that had been established in the capitalization process of HONDUTEL, in order to guarantee the new operator a rate structure that would allow it to recover its investment and, periodically reduce the cross-subsidy between international LD service and the other services. However, since the process failed, these contractual conditions never went into effect, and this provision was not applied (Tábora, 2007).

52 The idea of this plan was for the full rate to the USA to be US$0.42/min by the end of 2004, so that at the time of the opening up the rate should be US$0.28/min, thus making the impact on the market of the opening up less than it would have been at the time. It was not possible to obtain an official explanation for the reason why this plan was not continued, but it may be understood that the failure to apply the plan was influenced by the lack of a clear vision of the future of HONDUTEL, the entry into the market of sub-operators under the TpT programme and above all because of the difficulties of a fiscal nature being experienced by the current government (Tábora, 2007).

The high international LD rates undoubtedly contrast with local rates. As can be seen in Table 3.6, the cost of a 3 min call was $0.06, the lowest in the region after Costa Rica. It should be pointed out that the above data constitute an average, since in Honduras a distinction continued to be made between residential and business rates.53

HONDUTEL has introduced the system of prepayment in fixed telephony. In 2003, it launched Telecard on the market, which may be used in public telephones and in any terminal of its network. Telecard comes in denominations of 20, 60, 125 and 250 lempiras and may be used for local calls, national and international LD and communications to mobile telephones.

As we saw in previous sections, together with the privatization of INTEL, S.A. Panama granted the telecommunications company a period of exclusivity that was to last until January 1, 2003. Unfortunately the bidding process did not include definitions defining the rates each of them would charge. As González (2007) points out, it would have been logical, once the benchmark price for 49% of the shares had been set for the auction, for the process to favour the bidder who offered to charge the lowest prices, with precise quality requirements for the services to be offered.

However, the rates were negotiated after the concession contract had been concluded. Prior to the privatization, the cost of basic local telephony was a fixed charge of $10.50 with no additional variable cost. After privatization a fixed charge of $6.27 was established for the basic plan and a variable charge of $0.03/min. Table 3.6 shows the average costs of a local call. As can be seen, the prices in Panama were the highest in the region, with the sole exception of Mexico.54

Privatization with exclusivity also had as a consequence high prices in LD telephony. Between 1997 and 2002, a national LD call had a cost of $0.15/min. The opening up to competition reduced the price to $0.13/min in 2003, to $0.09/min in 2004 and was expected to fall in 2005 and 2006 to $0.07/min and $0.05/min, respectively (González, 2007). The costs of exclusivity are also reflected in international LD. The average prices of international calls excluding those to the USA reached a price of $1.86/min in 1998, $1.69/min in 1999 and 2000, and $1.57/min in 2001 and 2002. The opening up to competition evidenced how unjustified such high rates are. In fact, in 2003 the rates fell to an average of $0.40, in 2004 to an average of $0.25 and they were expected to fall to $0.12 in 2005. International calls to the USA cost $0.98 in 1998, $0.89 in 1999 and 2000 and

53 For details of the evolution of fixed rates in the period 2000–2004, see Tábora (2007).

54 The figures presented by González (2007) differ from those of the ITU and the World Bank. According to the aforementioned institutions, in 1997, the monthly charge for residential telephony was US$10.00 and in 2001 it was US$3.00. According to the World Bank, before and after the privatization the average cost of a local call remained at US$0.12 for 3 min, i.e. US$0.04/min. The difference may lie in the fact that the ITU figures for the monthly charge in residential telephony do not include business telephony, whereas the World Bank figures show an average of residential and business telephony.

$0.85 in the two years prior to the opening up. In 2003, the first year of competition, the rate dropped to $0.18, to $0.15 in 2004 and it was expected to reach $0.07 in 2005 (González, 2007).

As we saw before, in Costa Rica telecommunications are in the charge of ICE, a public enterprise that also provides electric power generation and distribution services. It is the responsibility of the Regulatory Authority for Public Services (ARESEP) to define the rates for services and supervise their quality. The legislation establishes that the services should be provided in accordance with their cost. Rates for telecommunications services are set under the rate system of the rate of return. In Costa Rica, this system translates into a rate that covers costs and the rate of return on capital (or development rate), understood as the income necessary for ICE to be able to replace or increase the assets it needs to provide its services. In operative terms, ARESEP applies the system of "price caps" to adjust ICE's rates. In order to guarantee that service quality is maintained, an adjustment factor for quality is included that affects only the component of surplus or return for the application of the formula, so that the resources to cover the costs of providing the service are guaranteed.

ARESEP has approved rate increases lower than those requested by ICE, with the purpose of inducing the firm to lower costs and increase productivity. The outcome of this policy during the period 1997–2005 has been a reduction in the basic residential rate equivalent to 27% (Vargas and Hernández, 2007). Moreover, as can be seen in Table 3.6, the average cost of a 3 min local call was $0.03 until 2002, dropping to $0.02 in 2003, thus making it the lowest average rate at that time among the countries considered in this study.

There is wide-ranging debate as to the causes of the Costa Rican peculiarities. A first argument maintains that as a public enterprise, ICE faces lower differential costs than a hypothetical private operator because they do not include the earnings or normal profitability of all economic activities (also known as retribution for the entrepreneurial factor or opportunity cost) because this kind of enterprise cannot generate profits. Also contributing to this situation would be the fact that its nature as a public enterprise exempts it from payment of income tax, other municipal taxes, as well as the right of way on public highways and use of the radio spectrum. As the only operator in the market it also enjoys significant economies of scale. Another feature that contributes to explaining the low rates stems from the synergies the enterprise has by also operating the electric power distribution networks, such as those derived from posts for stringing electrical, telephone and optical fibre cables. Cross-subsidies also play a role. As in the past in the majority of countries, residential telephony was subsidized by the higher rates for fixed telephony service to companies (business) and international telephony. The considerable reduction in rates for international telephony at world level, together with technological progress – which has made possible various forms of contracting this service from the country with companies other than ICE – has obliged the enterprise to seek alternative sources to finance cross-subsidies. Thus, at present mobile telephony has become the source of financial resources to subsidize less-profitable telecommunications services (Vargas and Hernández, 2007).

2. Evolution of Mobile Telephony Rates

An analysis of mobile telephony rates is substantially more difficult than in fixed telephony. What is involved, in general, is a market open to competition and therefore, normally, without the obligation to report rates. It is also a telecommunications service segmented between the post-payment and prepayment markets, often with sharply differentiated rates. Also, since there is considerable competition in this market, each one presents a wide variety of plans with many differences in nature, making it difficult for users to choose among them.

In the prepayment market, competition can take place in relation to prices per minute of airtime, as well as in connection with other variables such as: denomination of the card, duration of such denomination, grace period to recover available balances when its duration expires, and promotions and services offered by the prepayment system. Furthermore, the strategy followed by the licences has been to differentiate themselves to avoid open competition in the majority of the above variables (González, 2007).

In his analysis on competition in telecommunications in El Salvador, Argumedo (2007) draws four conclusions: (i) calling the USA is cheaper than communications between cellular phones in the country; (ii) a call from a cellular phone to a fixed line is more expensive than from cellular to cellular; (iii) calls between cellular phones of the same company are cheaper than between different companies; and (iv) calls from post-payment cellular phones are lower than those made by prepayment.

In Guatemala, the existence of greater competition among the operators has given consumer prices an interesting dynamic. In October 1999, Telefónica entered the market and immediately introduced the "caller pays" system and a 70% reduction in the price of outgoing calls. The entry of Telefónica also contributed to altering the characteristics of competition by diversifying rate plans, introducing post-payment or prepayment systems and offering individual and business (corporate) plans. By January 2005, 88% of mobile telephony was in the form of prepayment and only 12% in post-payment (Urízar, 2007).

In Honduras, the concession contracts of the two mobile operators CELTEL and MEGATEL established rate ceilings for mobile telephony services. Three charges were identified: the basic monthly rate which was set at the equivalent of $30, the charge at full rate set at $0.25/min and the charge at reduced rate set at $0.18/min. These rates set first for CELTEL were to last ten years (until 2006). Operators had the freedom to determine the duration of the different rates and even the possibility of introducing a super-low rate. Even though nominal rates have remained fairly stable in an analysis of consumption per subscriber, in real terms rates went down 31% in 2003–2005. Additionally, on average, during the latter period users used this service approximately 44% more than the time they talked in 2002 and for the same amount of money (Tábora, 2007).

In the prepayment market, CELTEL offered cards in seven denominations, between 25 ($1.31) and 450 ($23.68) lempiras, which are differentiated by the number of minutes and the duration but have the same cost per minute–$0.2632. In the post-payment market, CELTEL offered five plans that include between 60 and 400 min. The prices per minute varied from $0.25 55 (Plan Móvil 15–60 min for a total price of $15) to $0.19 (Plan Móvil 75–400 min for a total price of $75). MEGATEL, Honduras's other mobile telephony operator, had six post-payment plans and their prices per minute ranged between $0.27 (Plan Alo 12–44 min for a total cost of $12) and $0.19 (Plan Alo–525 min for a total cost of $100).56

In Panama's case the companies compete first, in the card denominations. Whereas Movistar has cards worth US$2, 5, 10, 20 and 40, Cable & Wireless has US$5, 10, 15, 30 and 50.57 The price per minute also varies. In Movistar's case, the company had only two rates. For the US$2 card, the price per minute was $0.49. For the rest of the cards the price was $0.43/min. Cable & Wireless, for its part, has a greater variety of rates per minute: US$0.40, 0.34 and 0.32 for the US$5, 10 and 15 cards, respectively, and US$0.25 for the US$30 and 50 cards. Naturally, in the card denominations in which the two companies coincide, users receive a different amount of minutes.

In the case of the US$5 and 10 cards, the licencees do not compete with regard to the duration, as both of them last between ten and 30 days, respectively; nor do they compete with regard to the number of grace days for recovering the balance: their cards offer 15 days to recover balances. Market promotions have adopted the means to double, triple, etc. the value in minutes of each card denomination. According to González (2007), this is the most visible way in which both licencees compete in Panama.

In Costa Rica, as we saw in the preceding section, mobile telephony has replaced national and international LD telephony as a source to finance for other telecommunications segments, particularly residential fixed telephony. Mobile telephony, despite its lesser coverage, contributes more than half of ICE's total revenues for telecommunications services. These circumstances may of course explain the relatively low coverage of mobile telephony in Costa Rica compared both with the levels attained by other countries and with the coverage of fixed telephony (see Section IV).

55 An interesting detail in the prepayment form is that the cards already include 12% sales tax (ISV), and therefore the real nominal rate is US$0.235, an amount lower than the reference rates for the post-payment plans Móvil 15 and Móvil 20, which do not include ISV. This analysis shows that the prepayment rate (US$0.235/min) is less than the post-payment rate of the Móvil 15 (US$0.28/min) and Móvil 20 (US$0.2635/min) plans (Tábora, 2007).

56 In addition to the traditional plans, MEGATEL offers the "Control ALO" Plan, whereby a credit limit on consumption during the month is established (Basic Set Rate) and once the assigned credit limit is reached, the user has the option of entering an Alo Card to continue using the service. With this mechanism subscribers can ensure that their monthly bill will always be for the same amount, with the proviso that any excess will have to be paid in cash by means of prepayment. One of the main advantages offered by this plan is that it accumulates any unconsumed minutes for the following month, including both the minutes assigned and the minutes entered with an Alo card (Tábora, 2007).

57 The information corresponds to April 18, 2005 (González, 2007).

VI. Conclusions and Policy Recommendations

Privatizations and openness of fixed telecommunications promised a world of many operators, a substantive increase in quality and coverage of telecommunications services, as well as a drop in prices. The coverage and quality of these services have improved, but their expansion has been slower than expected, and even Costa Rica – a country which has persevered in keeping these services under the public ownership – has better coverage and prices than the rest of the nations included in this study.

Two central elements help understanding this situation. First, private monopolies have shown a remarkable strength to resist competition. Second, the usual shortcomings of sectoral regulation are greater in smaller economies, particularly in developing ones where two basic institutions – law enforcement and the markets – are weak.

In this context, to discipline the incumbent is quite difficult and high prices limit the population's possibilities to use this service. This is specially so for low-income sectors. Together with the privatization and openness path followed, the fixed telephony problems are associated with an institutional underdevelopment.

The different paths taken by the privatization and liberalization process of fixed telecommunications were a promise of a world with many operators, a much greater coverage and better quality of the service, as well as lower tariffs, especially as regards the granting or not of exclusivity periods, which seem to translate into significant differences when the intensity of competition and the prices of telecommunications services are analysed. The case of El Salvador shows that the opening up to competition in fixed telephony, without a period of exclusivity, has translated into a much higher participation of companies other than the incumbent. By contrast, the exclusivity period allows the consolidation of a non-competitive model which tends to become more pronounced due to institutional weaknesses that allow anti-competitive behaviour by the incumbent, especially during the period of opening up to competition.

Separate privatization of the fixed company and the mobile company appears to constitute a measure of singular importance in promoting competition. In fact, the application of this policy in El Salvador has resulted, despite the country's small size, in the presence of four mobile operators with well-balanced market shares.

The formal presence or absence of a competition agency is not sufficient to make a difference. In fact, in countries which have a competition agency with a history close to, or greater than, ten years (Mexico and Panama), the predominance of the incumbent company does not differ substantially from what happens in countries without a competition authority. This seems to be associated with the weakness of the competition agency.

The mobile telephony represents a very different situation. Its technological characteristics and an active public competition policy made the participation of many operators possible. The intense competition in this industry had an impact on prices and promoted innovative commercialization techniques, allowing very low-income customers to participate in this market. Prepayment reduced the access cost to this service and the business model followed by this industry focused on massive use rather than monopolistic prices to raise its profits. From the users' point of view, the prepayment mechanism is helpful, because expenditure is always under their control and when they lack resources to make a call, they can always receive them. Many transaction costs are eliminated for the enterprise and its customers. In this way, without the need of public subsidies, mobile telecommunications is making headway towards universal access to voice communication services.

Within the net and services convergence framework there is the possibility that the strong competition in mobile telephony may be transmitted to the rest of the industry. This would provide important benefits for consumers and improve firms' competitiveness. Nevertheless, high entry barriers together with weak competition regulation may reinforce the increasing importance of regional duopolies in telecommunications, such as Telefónica y TELMEX – América Móvil, and offset such positive tendency mentioned above.

Tariff regulation has to make sure that competitive market pressures are simulated for the fixed telephony incumbent, if not naturally existent. Price reductions to near competitive levels can increase/generate substantive coverage. This is particularly important in fixed telecommunications because it provides private access to the wide band. This is particularly true since fixed telecommunications still offer important advantages over mobile telecommunications for these purposes.58 Independently from the arguments that favour the total liberalization of this sector, sectorial agencies need to regulate this sector in the best possible way in the medium run.

Regulation, though, is not enough. Competition promotion is increasingly important. It is particularly relevant to eliminate barriers to entry permanently (deregulation and surveillance of interconnection blocking or overpricing, among other anti-competitive behaviour), to control mergers as well as other measures to ensure price competition. For these measures to succeed, four elements are crucial: a strong competition agency, a strong coordination between the latter and sector regulation, improvement of the judiciary system performance in these matters and the development of a supranational competition legal system.

In Costa Rica's case, the future of telecommunications is linked to the more global decisions the country takes after the presidential elections of February 2006. The fundamental decision is undoubtedly the one relative to ratification of DRCAFTA. If approved, Costa Rica should approve a General Telecommunications Law, create a regulatory agency for the sector and open up Internet, private networks services and wireless telephony to competition.

58 Mobile telephony is a very good option for voice transmission, but in the short run it is not a comparable competitive alternative for Internet for the population at large. This will probably change in the longer run.

The free trade agreement does not include the obligation to proceed to the privatization of ICE. Nevertheless, those in favour of maintaining its public nature fear that with the opening up the future of the enterprise may be threatened definitively by competition from international operators, particularly Telefónica España and the TELMEX – América Móvil group.

It is in this context that large telecommunications firms are readjusting their organization structures. As in the rest of the world, there are clear signals that fixed telephony has ceased to be the nucleus of the operation, as it has lost importance as the main source of income and its traditional monopoly over the last mile has been weakened.

Towards the end of the 1990s, Telefónica took two important decisions: first, to acquire the shares of the retail owners of its affiliates, and second, to structure their operations according to the different telecommunications segments. The initiative made it possible to better organize the main lines of business, but somehow it ran counter to the trends defined by the process of convergence.

Users' growing interest in relying on providers that can supply a comprehensive package of telecommunications services in some way raises doubts about the above-mentioned organizational model. This is translating into the search for more efficient forms of organization59 and into a systematic policy of incursion into fixed telephony. Thus, in the case of Panama, Telefónica Móvil has requested a concession for the provision of fixed telephony, and in Mexico the company has established a strategic alliance with Avantel and has also indicated that they are seeking the acquisition of a fixed operator.

One aspect that is taking on growing importance is the debate regarding the best means of treating the industry. Even though an analysis of the different markets into which telecommunications are divided continues to be useful, the way in which the business is structured, the multi-purpose nature of the networks as a result of convergence and the companies' own operation make it necessary to move on from isolated treatment of each segment towards comprehensive treatment of the industry. The formerly different networks are becoming integrated so as to build a single platform for the provision of fixed and mobile telephony, LD, data transmission and television services. Companies use all their assets to compete in the markets for each of the end services.

Furthermore, the companies that divided up just over five-years ago according to the traditional segments, fixed and mobile telephony and Internet, have begun a process of reintegration and are seeking new forms of internal structuring. Though this tendency seems essential to make the best use of new technologies, regulation and competition institutions must prevent anti-competitive practices.

59 Among them being the introduction of de facto organizational formulas in which the different companies converge, with the establishment of directorates of business areas, which are served by the network and planning boards of directors, etc.

Bibliography

Ansorena, C. (2007)s, Competencia y regulación en las telecomunicaciones: el caso de Nicaragua (LC/MEX/L.785), serie Estudios y Perspectivas No. 73, ECLAC/IDRC, July.

Argumedo, P. (2007), Competencia y regulación en las telecomunicaciones: El caso de El Salvador, serie Estudios y Perspectivas N° 73, ECLAC/IDRC, Mexico, February.

Estache, A., M. Manacorda and T. M. Valletti (2002), Telecommunications reform, access regulation, and Internet adoption in Latin America, Economía, vol. 2, N° 2, Spring 2002, pp. 153–217.

FUSADES (Fundación Salvadoreña para el Desarrollo Económico y Social) (1998), "Las Telecomunicaciones en El Salvador: un sector estratégico para el nuevo siglo", Boletín Económico y Social, No 153, August.

González, R. (2007), Competencia y regulación de las telecomunicaciones: el caso de Panamá, serie Estudios y Perspectivas N° 74, ECLAC/IDRC, Mexico, February.

International Telecommunication Union (2005). http://www.ine.int/ITU-D/icteye/reports.aspx#.

Mariscal, J. and E. Rivera (2005a), New trends in the Latin American telecommunications market: Telefonica & Telmex, Telecommunications Policy, N° 29, pp. 757–777.

Mariscal, J. and E. Rivera (2005b), Organización, industrial y competencia en las telecomunicaciones en América Latina: las estrategias empresariales, serie Desarrollo Productivo N° 169 (LC/L.2423-P), ECLAC, Santiago, Chile.

Mariscal, J. and E. Rivera (2007), Regulación y competencia en las telecomunicaciones mexicanas, serie Estudios y Perspectivas N° 83, ECLAC/IDRC, June, Mexico.

OECD (2007), Communications Outlook 2007. http://213.253.134.43/oecd/pdfs/browseit/9307021E.PDF

Ramírez, F. (2005), Interconexión de redes telefónicas en México: solución de disputas y búsqueda de consensos, Mexico, unpublished.

Rivera, E. (2006), Regulación y competencia de las telecomunicaciones en Centroamérica: un análisis comparativo, in C. Schatan and M. Avalos (Coords.), Condiciones y políticas de competencia. Economías pequeñas de Centroamérica y el Caribe, FCE y ECLAC, Mexico.

Tábora, M. R. (2007), Regulación y competencia en el sector de las telecomunicaciones en Honduras (LC/MEX/L.781), serie Estudios y Perspectivas No. 84, ECLAC/IDRC, México, June.

Urízar, C. (2007), Competencia y regulación de las telecomunicaciones: el caso de Guatemala, serie Estudios y Perspectivas N° 75, ECLAC/IDRC, Mexico. February.

Vargas, L. and G. Hernández (2007), Competencia y regulación en las telecomunicaciones: el caso de Costa Rica, ECLAC/IDRC, Mexico, forthcoming.

Waltsten, S. (2000), "Telecommunications privatization in developing countries: the real effects of exclusivity periods". http/www.isnie.org/ISNIE01/Papers01/waltsten.Pdf.

Chapter 4
Competition and Regulation in the Banking
Systems of Central America and Mexico:
A Comparative Study

Eugenio Rivera and Adolfo Rodríguez

Introduction

Competition in the financial sector, especially in the banking sphere, has been a very polemic topic. The adverse selection and moral risk that characterize this sector explain the high regulation of its economic activities. This feature is geared towards solving the important information problems faced by the economic actors, so as to protect savers' interests and ensure the stability of the financial sector as a whole. This objective, nevertheless, contradicts to some extent the aims of competition policy. On the one hand, excessive competition may create difficulties for the bank supervisor: it is easier to control a few large banks than having to do so with a large number of small ones. On the other hand, an excessive competitive dynamic may force banks to transfer a large part of their surplus to the consumer, limiting the banks' capitalization capacity and may lead some of them to undertake excessive risks to face competition. This explains to a certain extent the existence of some regulatory arrangements – such as the requirement of relatively high minimum capital – that promote bank concentration. But, at the same time, the absence of competition can weaken market discipline and allow for inefficient practices, while the excessive concentration can reduce the supervisor's relative power as he faces banks that are "too big to fail".

However, the recent evolution of financial markets tends to solve this contradiction. The development of banking technology has created, more than any other economic sector, a competitive worldwide space, so that local banks are facing, more than ever, a strong competition. This cannot be avoided by measures that promote local bank concentration; on the contrary, concentration can worsen some of the negative effects of excessive competition on banks' behaviour, particularly in the negligent handling of risk. Competition promotion in local markets, together with measures that improve the local bank competitiveness so as to close the gap with the international one, is the best complement to the prudential norms. These should be revised so as to remove aspects that promote concentration and lack of competition in this sector.

Understanding banks' competitive strategies and their performance requires an analysis of the institutional conditions in which financial institutions operate. Such conditions include: (i) the regulatory framework, which includes prudential regulations, monetary policy and taxation; (ii) the characteristics of the region's financial infrastructure – especially those concerning liquidity management and payment systems; (iii) international connectivity; and (iv) non-bank sources of contestability in the banking market, particularly public debt and stock markets. This is the context that determines banks' competitive strategies in the region.

In fact, in this context, the Central American banks have developed a set of strategies to ensure their continued existence in the setting of the industry's growing internationalization and to respond to the demands imposed by the application of the Basel standards. The following feature among those strategies: (i) a process of modernization and efficient organization of activities; (ii) development of offshore banks in parallel to local ones; (iii) an active policy of mergers and takeovers which has reduced the number of banks and increased concentration indicators of the five largest institutions; (iv) regionalization of activities and of the ownership of financial groups; and (v) growing involvement in financial activities formerly barred to the banks.

With this context in mind, the Section I analyses the background and salient features of the history of the banking system in the different countries of the region, as well as the main characteristics of the institutional framework governing regulation and promotion of competition. Section II consists of a study of the main determining factors for the banks' competitive strategies. Section III analyses the main strategies developed by bank operators. Section IV shows certain indicators regarding the banking system's performance from the point of view of both management efficiency and intensity of competition in the system. The study ends with the main conclusions drawn and some policy recommendations in Section V.

I. Organization of the Banking System and Institutional Framework for Regulating and Promoting Competition in Central America and Mexico

1. Historical Background of the Banking System in the Region

Although the bank's historic trajectories in the different nations are extremely varied, they show at least three similarities: the presence of a strong public bank at some point, the generalized tendency to privatization of such bank (with the exception of Costa Rica) and the presence of political or economic turmoil which have threatened the industry's financial solidity.

Recent trends in the banking sector reveal a modernization process and an improvement in its regulation, even if at a different pace in each country. There has been a considerable progress in the prudential regulation application and different steps have been taken in order to achieve a consolidated supervision of the whole financial sector. The authorities' concern for financial stability has limited their effort to promote more competition in these markets. This can be appreciated in the fact that regulatory bodies have not included competition as one of their most important tasks, as well as in the absence of a competition agency, or else in their judicial limitations to undertake these cases.

Over the past 15 years the banking system in Central America and Mexico has undergone major transformations. A brief look at the development of the banking system in Mexico reveals a traumatic track record. In the context of the international-debt crisis in the early 1980s, the Mexican government announced the nationalization of the banking system. Less than ten years later, in May 1990, the Mexican government, then headed by President Carlos Salinas de Gortari, announced the constitutional amendment that made privatization possible. The government's aim was to maximize public revenues. The privatization process was structured as a series of successive auctions in which the losers could participate in bidding processes for subsequent blocks of shares. The process was open not only to groups with banking experience, but also to groups linked to the stock exchange. No kind of assessed value of assets or minimum prices was made known, so the sale was carried out at an average of three and a half times the book value. The high sale prices were subsequently considered to be one of the reasons behind the collapse of most of the banks in the context of the 1994–1995 "Tequila crisis". Several causes were adduced to explain the above-mentioned crisis, among them the fragility of the Mexican banking system, stemming from the low quality of the banks' credit portfolio and the high level of leverage in the corporate sector. These circumstances followed from the rapid expansion of credit and the overvalued prices of assets and guarantees which took place during the years of economic expansion in Carlos Salinas' term of office and were associated with the weakness of prudential regulation. One particularly serious situation was the growth of related credits without adequate backing. Although less than eight years had elapsed between nationalization and privatization, this time it took four years for the privatized banking system to collapse. The tequila crisis brought in its wake high delinquency due to high interest rates, contraction of the supply of uncommitted funds, reduction in the level of financial intermediation and economic slowdown, which made it necessary for the state to intervene in the capitalization of banking institutions. In 1999, only 35% of the 20 banks in operation came from the privatizations; two of these – Bancrecer and Inverlat – were controlled by Institute for the Protection of Bank Savings (IPAB), another two had been taken over by foreign banks (servicios financieros integrados (SERFIN) by Santander Mexicano and Banco de Comercio (BANCOMER) by Banco Bilbao Vizcaya Argentaria (BBVA) ) and Banco Internacional (BITAL) was being capitalized by European banks. Thus, by 1999 only two of these institutions – BANAMEX and Banorte – had retained 100% Mexican stock ownership (see Avalos and Hernández, this volume).

As in Mexico's case, the commercial banks and savings and loan associations of El Salvador were nationalized in 1980.1 Ten years later, in 1990, the process was

1 Legislative Decree No. 158: "Law on the Nationalization of Credit Institutions and Savings and Loan Associations" dated March 7, 1980.

reversed and the government set in motion a process to privatize the sector. This evolution has taken place together with several legal changes, among which are the Organic Law of the Central Reserve Bank of El Salvador, passed in 1991, which included a prohibition on directly or indirectly financing the state (Article 74) and eliminated the authority of that institution to set the exchange rate and interest rates. These have been subject to market forces ever since. Probably the most important subsequent event in the development of the financial system took place on January 1, 2001, when the Law on Monetary Integration went into effect. The law introduced the use of the US dollar as legal tender and stipulated that all banks and financial operations be denominated in that currency. Naturally, the role of the Central Reserve Bank had to be redefined, particularly as regards monetary policy. In 2006, there were ten private banks and two state banks (Herrera, 2007).2

Nicaragua has not been free of vicissitudes. The nationalization of the banking system after the Sandinista revolution was followed by an initial period during which a state banking system was in operation (1980–1990); the next decade was characterized by a private banking system which gained a leading role in the country but underwent a severe crisis (2000–2002) paving the way for a process of concentration between 2003 and 2005 that had left the country with six banking institutions by 2006. Despite the ups and downs in the history of Nicaragua's banking system, or perhaps because of these, Nicaraguan financial capital has a strong presence in the region.

Undoubtedly the characteristic feature of the banking system in Costa Rica is a strong state sector with origins in the 1949 bank nationalization, which also provided for a monopoly on deposits based on the argument that the public's savings should serve the public interest and not private profit-making. Only one bank was not nationalized, Banco Lyon, which specialized in international operations to support exports. Nevertheless, protected by a series of reforms, the private financial sector started attracting funds even since the end of the 1960s. From 1980 onwards, the private banks started having a wider space. Further leeway was afforded to private banks by authorizing access to Central Bank credit programmes and it was allowed to attract savings from the public at progressively shorter terms.

Panama stands out for being an international financial district. The modern Panamanian banking system has its origins in Cabinet Decree 238 of 1970 which established the National Banking Commission, the first attempt at state regulation of banking operations in the country. The Law formalized the presence of international banks in Panama by authorizing both local and offshore operations, depending on the interests of each institution. Banks that engaged exclusively in offshore transactions, however, could make local interbank placements, thus creating a highly active interbank market that forms the basis of the system to this day. The Bank Law gave every

2 Public banking in El Salvador has not been devoid of problems, an example being Banco de Fomento Agropecuario (BFA), which was the scenario of one of the financial system's biggest frauds, in which BFA lost approximately 24.8 million colones (US$2.84 million). Another negative aspect affecting public banking has been a poor-quality loan portfolio. At the end of 2002, Banco Hipotecario registered a 9.3% default rate and BFA 28.6% (Herrera, 2007).

bank the freedom to establish its own borrowing and lending rates and the granting of credit – with a few minor exceptions – was not restricted by general ceilings or by specific portfolio allocation by sectors. This rate-setting freedom for each bank paved the way for the integration of the banking system with the rest of the world, and the London Interbank Offered Rate (LIBOR) is the base for establishing the cost of credits. The number of banks increased steadily between 1970 and 1983. Notwithstanding, since the latter year, some important banks such as First Chicago, Libra Bank and Bank of America, among other smaller ones, began to shift their operations to other financial centres, as their presence in Panama was founded on the accounting records of sovereign loans to Latin America; but when those loans became delinquent, it was more advantageous to register them in centres where those losses could serve as a tax shield. Even so, many foreign banks continued to carry out operations from Panama and to cater to the local market (Fernández, 2006).

In Guatemala's case, a number of problems hindered approval of the new regulatory framework for the banking system throughout the 1990s. It was not until 2002 that the Monetary Law, the Law on Free Negotiation of Foreign Currency, the Organic Law of the Bank of Guatemala, the Law on Banks and Financial Groups, the Financial Supervision Law and the Law on the Stock and Commodities Market were passed. The delay in implementing the reforms has been associated with the outbreak of the 1998 financial crisis,3 which pointed out deficiencies and poor financial and administrative management within the framework of legislation that proved obsolete in the face of the financial innovations that had taken place (Balsells, 2007).

2. Institutional Regulatory Framework and Promotion of Competition

The modernization of the banking system, its growing internationalization, bank disintermediation processes and the importance of competition in ensuring the industry's efficient performance raise a series of institutional problems of prime academic and public-policy interest. This evolution, together with the intensification of prudential regulation, has demanded the development of greater competence from regulatory agencies. Domestic banks have grown in size and in their international links, while international banks, because of their large size and its global way of operating, pose a challenge for regulators mostly operating on a domestic basis. Hence, international cooperation among regulators takes on vital importance.

3 The 1998 crisis was brought on by risky investments in agro-export companies linked to traditional products such as coffee and cardamom, which suffered marked price drops that year. In addition, important banks invested in securities affected by the "Asian crisis", including Russian government bonds that fell into a non-payment situation. International problems, together with the coffee crisis and the appearance of natural phenomena such as Hurricane Mitch, provoked a liquidity crisis in the banking system which rapidly raised the country-risk situation and led to the bankruptcy of a considerable number of financial firms (Balsells, 2007).

An intensely debated issue is the independence of the institutional regulatory framework from the government. At one end, the appointment of regulatory authorities by the Congress is seen as a guarantee for these authorities' autonomy, while at the other end this autonomy is seen as a potential source of a lack of coordination on economic policy and a kind of "total political irresponsibility". As we shall see, in most of the countries studied Executive Branch authorities intervene in the appointment and dismissal of regulatory authorities, and adequate mechanisms for safeguarding the independence of regulators are non-existent. The problem is compounded when the sector being regulated has a say in the designation of authorities and in approving the principal measures, an example being the Monetary Board in Guatemala.

Another institutional issue raised with growing insistence is competition. As we will see, recent developments have given rise to a process of mergers and takeovers which has produced a concentration of ownership and a regionalization of the banks. This process has taken place in parallel with a certain reduction of intermediation margins and commissions which nevertheless remain at relatively high levels, indicating a low degree of competition in the sector. This justifies discussion on the need to promote the role of competition agencies. Until late 2004, four countries in the region lacked a competition supervisor, and in the other three the effective powers of such agencies in the banking sector were highly limited or non-existent. Bank supervision agencies fail to appreciate the importance of competition and the involvement of competition agencies; in point of fact, bank supervisors still tend to consider banking as a special industry which on account of its importance to the overall performance of the economy should be protected from the "excesses" of competition regulation. The competition laws recently passed in El Salvador and Honduras acknowledge the importance of competition in the financial sector and establish that the competition agency should become involved in that sector just as in any other economic activity,4 for which purpose the means of coordination with financial supervision authorities should be appraised.

In Mexico the chief supervisor is the Ministry of Finance and Public Credit (SHCP), which authorizes the operation of financial groups, banks, insurance companies, pension fund administrators (Afores) and stockbrokerage firms and follows up on these entities through four commissions in charge of regulation and prudential supervision: the National Banking and Securities Commission, the National Insurance and Sureties Commission, the National Commission on the Retirement Savings System and the National Commission for the Protection and Defense of Users of Financial Services. Analysts maintain that the Mexican financial system is dominated by the major financial groups which own banks, stockbrokerage firms, insurance companies and Afores, complicating regulation due to the overlapping of activities engaged in by each entity, the existence of cross-subsidies not properly registered in accounting ledgers and coordination difficulties among the regulatory agencies.

4 In Mexico's case the powers of the Federal Commission on Competition (CFC) apply to the financial sector in the same way as to any other sector. Nonetheless, the CFC's actions have focused mainly on reviewing mergers and takeovers (Avalos and Hernández, this volume).

In contrast to other countries of the region, since 1993, Mexico has had a competition agency, the Federal Commission on Competition, which is in charge of supervising mergers in the industry. Despite the modernization of the regulatory agencies, one problem that has affected regulation is the lack of modernization of the Judiciary.5

The Salvadoran financial system is governed by four institutions: the Superintendence for the Financial System (SSF), in charge of supervising and regulating the activities of commercial banks, insurance companies, bonded warehouses, bureaux de change and other concerns related to financial activities; the Securities Superintendence, in charge of supervising securities intermediaries, central securities depositories and the stock exchange; the Pensions Superintendence, responsible for supervising and regulating Pension Fund Administrators; and the Central Reserve Bank, which is in charge of promoting and maintaining monetary stability and developing an efficient, competitive financial system. The SSF forms part of the Central Reserve Bank of El Salvador and possesses autonomy in administrative and budgetary matters and in exercising the powers conferred on it by the Law. The SSF's management is headed by a Directive Board and the Superintendent, the latter nominated by the council of ministers and approved by the President of the Republic. The Bank Law establishes that the SSF is in charge of authorizing bank mergers and such mergers must be carried out in accordance with the rules established by the Code of Commerce. Throughout the period studied, El Salvador has lacked a competition agency. The Law on Competition was passed in December, 2004, and the agency began operating in January, 2006 (Herrera, 2007).

In Guatemala, the law stipulates that supervision of the entire financial system, including banks, is incumbent upon the Superintendence of Banks, a Central Bank agency that acts under the general directorate of the Monetary Board. The latter is responsible for submitting a list of three candidates from which the President of the Republic appoints the Superintendent, whose resolutions in regard to oversight and inspection may be appealed before the Monetary Board. The Monetary Board grants or denies authorization for the incorporation of banks and foreign bank branches, and for bank mergers. The Monetary Board's powers and composition translate into a low level of autonomy for the supervisory body both from the government and from the private sector itself (Balsells, 2007).6 Guatemala had neither competition law nor agency yet in 2006.

Nicaragua also has a system with a sole regulatory agency. The Superintendence's higher organs are a directive board, a superintendent and a vice-superintendent.

5 According to a study, 60% of mercantile executory processes do not go beyond the first stage of the proceedings, i.e. summons and attachment. Moreover, of the remaining 40% of cases, only half reach a final decision, in other words, of every 100 complaints admitted, only 20 reach a verdict (Avalos and Hernández, this volume).

6 The board is chaired by an official appointed by the President of the Republic for a four-year term who is also chairman of the Bank of Guatemala (BANGUAT), and its members include the Ministers of Public Finance, Economy and Agriculture, a member elected by Congress, a member elected by trade, industry and agriculture business associations, a member elected by the National Banking Association and a member elected by the Higher Council of the University of San Carlos of Guatemala.

It authorizes, regulates and supervises the operations of the following entities: banks, finance companies, insurance companies, the stock exchange, seats on the stock exchange, general bonded warehouses, leasing companies and Tourism Investment Capital Funds (FONCITURs). Its sphere of duties makes no reference to considerations of concentration, dominant position or abuse of market power and their implications for competition; the criteria for approval of mergers or takeovers contain only aspects that denote concern over the system's solvency. In accordance with the current provisions of the Bank Law, by virtue of bank secrecy in Nicaragua that sector remains outside the jurisdiction of any future Competition Agency (Ansorena, 2007).

In Honduras, the Law of the National Banking and Insurance Commission (CNBS) was passed, becoming a de-concentrated agency of the Presidency of the Republic, assigned to the Central Bank of Honduras (BCH), but with absolute technical, administrative and budgetary independence. Its mission is to supervise and regulate but the law does not assign any competition supervision function to it (Tábora, 2007).

II. Determining Factors in the Banks' Competitive Strategy

In the struggle to secure market share and an adequate rate of return, the banks try out different strategies that are not only determined by their objectives and goals, but also to a large extent constitute reactive behaviour in the face of the external conditions in which competition operates. The determining factors of these external conditions could be classified into four groups, mainly of institutional origin and relatively independent of the banks: (i) the regulatory framework; (ii) the infrastructure of the financial sector; (iii) the connectivity of the financial system; and (iv) the internal sources of contestability. This section discusses these factors, their effects on banks' competitive strategies and their results.

1. The Regulatory Framework

The banks face a regulatory framework that operates in different fields and has consequences on their performance and on the way competition operates in each country and in the region as a whole. One could hardly explain the course of the banking system's development in the different countries without taking that framework and its vicissitudes into account.

1.1. Discriminatory Barriers to Entry

The regulatory provisions with the most obvious effects on competition are those concerning entry barriers. There are barriers that hinder the entry of new banks due to existing banks' "natural" advantages related, for example, to economies of scale or relationship lending. Other barriers stemming from the regulatory framework indistinctly affect all banks seeking to establish themselves in a particular market, such as minimum capital requirements, for example. Finally, there are barriers that do not equally affect all the banks seeking to establish themselves in the market, but that discriminate on two grounds: origin of capital (foreign or domestic) and ownership of capital (public or private). These are known as "discriminatory entry barriers".

At present there are no absolute barriers to the entry of particular banks, although they have existed in the past. The most extreme case is Costa Rica which, as we saw in the preceding section, nationalized all the existing banks and implemented a state monopoly on deposits from the public in 1949; this was followed by a lengthy period during which for all intents and purposes only state-owned banks existed. This monopoly began to be loosened in the mid-1980s when private banks were authorized to take deposits from the public for terms longer than six months and was eliminated in 1995.7 The extensive network of branches developed by the public banks and the existence of an unlimited state guarantee on their deposits has made Costa Rica the only country in the region in which public banking clearly predominates.

The opposite extreme is exemplified by Panama, which imposes no restrictions on the banks to establish in its territory. The 1970 Bank Law authorized operations, from offices established in Panama, carried out, completed, and having effects in other countries, as well as the establishment of bank agencies in Panama. This set of provisions, complemented by a deregulation process that included the liberalization of interest rates and later the elimination of reserve requirements, gave rise to the development of the "financial district" which became an important offshore destination for the region's banks and for some of the world's largest banks.

Between those two extremes there are various degrees of discriminatory entry barriers in the region. With regard to ownership of capital, the barriers that discriminate private banks as against public ones concern the obligation of private institutions to manage their funds through public banks, income tax exemptions for public banks and the existence of a state guarantee on deposits in public banks, which in countries such as Costa Rica and Guatemala play an important role for the public. By contrast, the law in some countries stipulates a number of specific uses for public bank profits and the legislation imposes a series of restrictions on them regarding employment and administrative hiring.

As to the origin of capital, no country in the region has restrictions on the establishment of foreign banks, although in some of them there are certain limitations on the establishment of offices or branches: only banks subject to local regulation and supervision can operate in the national territory. This restriction eliminates the

7 In exchange for eliminating the state monopoly on deposits, Costa Rican legislators provided that private banks taking sight deposits (current accounts) should place 17% of their acceptances in public banks so that the latter funnelled them into development credit.

possibility of banks established abroad to openly offer their services in the country, although they do so surreptitiously, as will be seen on analysing the degree of connectivity in the region and the role of offshore banking.

1.2. Prudential Regulation

Other than El Salvador and Panama, all the countries of the region have undergone bank crises (Mexico, Nicaragua) or periods of serious bank stress that have included bankruptcies and interventions (Costa Rica, Guatemala, Honduras). This has paved the way for considerable efforts to update prudential regulations and supervisory capacity. The countries of the region have gradually been adopting the prudential norms issued by the Basel Committee. These norms seek, among other aims: (i) to define minimum conditions for the establishment of a bank as regards its capital and the honour of its shareholders; (ii) to adjust each bank's capital base to its risks as a guarantee for absorbing unforeseen potential losses; (iii) to establish mechanisms to assess asset quality, risk and coverage by creating reserve funds; (iv) to set limits on portfolio concentration and restrictions on extending credit to persons linked to the bank; (v) to foster procedures for assessment and comprehensive management of other sources of risk, such as country, infrastructure and market risks; and (vi) to regulate the use of internal controls and client assessment systems to prevent money laundering.

Among the Basel standards that most influence competition in the banking sector are those on capital requirements, which have gradually been increasing since the 1990s. On the one hand, the minimum capital required to establish a bank was around US$8 million on average in 2006. This amount is far below the minimum capital required in OECD countries and even in other Latin American countries (US$25 million in Chile). And on the other, the compulsory capital adequacy ratio (CAR) is generally higher than that recommended by Basel, which is explained by the greater opaqueness in information and the greater risk associated with these markets. The compulsory ratio fluctuated in 2004 between 7.5% in Guatemala and 13% in Panama (see Table 4.3). This ease, however, can be partly explained by the liberality with which the credit portfolio tends to be classified, the quality of which would require provisions that would lower the effective CAR. In any event the gradual increase in capital requirements has played an active role in furthering mergers and takeovers since the beginning of the present decade.

Other Basel standards affecting competition are those concerning credit, particularly limits on portfolio concentration and credit to related persons. The main problems facing the banks in the region are associated with credit risk, since information on portfolio quality – debtors' capacity to pay and loan security – is frequently lacking, until the economic situation deteriorates and a high proportion of the portfolio becomes delinquent. The excessive concentration of the portfolio in a few sectors of the economy, as well as with loans to related persons, is the origin of most banking crises and interventions in the region.

Table 4.1 Central America and Mexico: capital requirements and reserve coefficients, 2005 (%) (Rodlauer and Schipke, 2005)

a El Salvador and Panama are totally "dollarized".

1.3. Monetary and Exchange Policy

Monetary policy in most of the economies of the region has been marked by central bank losses and monetary policy. Aside from Panama, officially dollarized since 1904, all the countries had fixed dollar exchange rates which were abandoned during the course of the 1980s as a result of inflationary processes, armed conflicts or financial turbulence: Nicaragua abandoned the fixed exchange rate in 1979, Costa Rica in 1981, El Salvador in 1983, Guatemala in 1984, Honduras in 1990 and Mexico in 1994. Abandonment of the fixed exchange rate often took place after an accumulation of imbalances which provoked a free fall, after which various exchange systems were adopted: following a relatively short period, Costa Rica, Honduras, Guatemala and El Salvador adopted a crawling peg, generally adjusted in line with expected inflation; later, El Salvador adopted dollarization as of 2001, and Guatemala an independently floating system since 2003. Mexico, for its part, has allowed its currency to float freely, although there have been episodes when the Central Bank has intervened forcefully in the foreign exchange market according to pre-established rules. In all cases, leeway in monetary policy is tightly restricted by exchange goals.

Moreover, the foreign exchange crises experienced by the Central American countries during the 1980s obliged the Central Banks to intervene in order to avoid the collapse of the commercial banks. This caused excess liquidity which had to be siphoned off through the placement of securities issued by the Central Bank itself, the service of which entails quasi-fiscal losses of several percentage points of gross domestic product (GDP) and creates chronic excess liquidity. The need to continually siphon off this excess liquidity and keep to foreign exchange goals has compromised the autonomy of monetary policy. This is aggravated by deficiencies in the payments infrastructure and in public-debt management, which has hindered the organization of interbank markets that enable the monetary authority to intervene through indirect instruments. Hence, the main monetary-policy instrument is cash reserves: aside from Panama, which lacks bank reserves, reserve coefficients fluctuate between 10% (Costa Rica) and 20–25% (El Salvador) (see Table 4.1). This way of implementing monetary policy has a high cost for banks, which transfer them to their clients through high intermediation margins (as cash reserves are not remunerated). This places local banks at a considerable disadvantage compared with their international competitors and motivates them to shift their operations elsewhere.

In Mexico the elimination of reserve requirements on new deposits by Banco de México coincided with a considerable net inflow of foreign capital just at the time of the re-privatization of the banks in 1991. This led to an increase in liquidity in the banking system and hence by a marked expansion of credit, also made possible by the weakness of the banks' supervision and risk management during the decade when they were nationalized. This expansion contributed to turning the 1994 exchange crisis into a serious solvency crisis.

Another consequence of the exchange and monetary environment is the growing informal dollarization experienced by these economies. The Central American Monetary Council (Consejo Monetarios Centroamericano, (CMCA) ) has estimated that in 2001, around US$1.401 billion was in circulation in the region, representing 44% of the monetary emission; in 2004 that dollar availability had risen to US$2.297 billion, 72% of monetary emission (CMCA, 2004).8 Between 1996 and 2003 dollar deposits in commercial banks – excluding Mexico and Panama – increased from 20.2% to 31.1% of overall deposits, and dollar credits from 15.7% to 28% of overall credit. This trend heightens the banking system's liquidity risk, insofar as there are less mechanisms for protection from the mismatching of maturities between dollar deposits and credits, as well as the credit risk, since a large proportion of debtors do not have dollar incomes and in the event of a devaluation many of them would become insolvent, which would compromise the health of the creditor banks' assets. This outcome of monetary and exchange policy, as well as the preceding one, has implications for competition in the region, particularly if not all the countries have prudential rules that make the price system reflect the risks inherent to informal dollarization.

2. Financial Infrastructure

The shortcomings suffered by practically all the countries of the region, except for Mexico, in the development of the financial infrastructure, have direct consequences on competition between local banks and their foreign competitors, as well as between large and smaller banks.9 In addition, they help to explain some of the

8 These estimates give an idea of the intensity of the dollarization process, although the amounts exclude Mexico and Panama and include the Dominican Republic.

9 In 1994, Banco de México undertook a comprehensive reform of payment systems which reduced the character risk and attained a high degree of security and operative reliability in settlement of payments.

competitive strategies that will be analysed in the next section. Two components of the infrastructure are especially relevant from the point of view of bank competition: interbank markets and interregional payment transfer systems.

2.1. Interbank Markets

Commercial banks manage their liquidity mainly in interbank markets by trading and borrowing public-debt securities. The existence of efficient interbank markets from the point of view of price-setting, but in particular of the timeliness with which operations are agreed and settled, is essential to them for optimal liquidity management. None of the countries of the region, with the exception of Mexico, have such markets. Two circumstances account for this.

First of all, central government-debt management has not provided, among its objectives, for the development of the public-debt market: non-standardized governments bonds are issued, there is no constant presence on the market, placements are made through mechanisms lacking in transparency – such as retail windows or unilateral negotiations with public or private institutions – and the organization of secondary and bond-execution markets has been left in private hands. The result is a practically non-negotiable stock of public debt which cannot be used, as in more developed countries, to manage banks' liquidity and risk and for the execution of monetary policy.10

Second, local payment systems stand as an additional hurdle to the organization and operation of interbank markets. Aside from Mexico, Guatemala and Costa Rica, no country in the region has payment systems that enable transfers between banks to be made in real time, which is a basic infrastructure for the development of an interbank market. Furthermore, in Guatemala and Costa Rica no measures have been taken to allow the use of real-time gross settlement (RTGS) infrastructure for an interbanking market to work. Therefore, the payment system lacks the mechanisms required to handle liquidity and is not used by banks for high value payments.11 This means that, except for Mexico, in order to transfer money, banks have to resort to two types of channels. On the one hand, clearing houses limit payment timeliness (because cheques are generally cleared in t + 1) and increase the systemic risk, as clearing houses lack risk management mechanisms for large payments; the funds a bank expects to receive are generally tied up in other operations, so that one bank's default can cause a chain of defaults. On the other hand, the channel used by banks to make their payments is through the Central Bank. Here a commercial bank can request the Central Bank, by telephone or fax, to debit its reserve account in order to credit another bank's account for the amount in question.

10 See World Bank and IMF (2000) and Litan et al. (2003), on the importance of an adequate management of the public debt the limitations faced by emerging markets.

11 See IMF (2005 and 2006), on the shortcoming of the regional payments systems, especially for its judicial framework.

This channel is too inefficient to allow the development of an active interbank market.

Beyond the consequences that the absence of public-debt interbank markets has for the state, since the lack of market liquidity increases the cost of public indebtedness and hinders the use of modern monetary-policy instruments, it distorts competition and limits local and small-bank competitiveness. In fact, the lack of mechanisms to handle liquidity forces commercial banks to maintain excessive liquidity at great financial cost in order to meet the vicissitudes of their daily needs and contract lines of credit with foreign banks to address their diverse contingencies. This increases intermediation costs and creates a dependence on international markets with regard to liquidity management and introduces competition distortions in the different banking systems. On the one hand, it gives foreign banks an advantage over local banks, since liquidity management is less costly for the former than for the latter. On the other, the large banks have an advantage over the smaller ones, since many of these cannot obtain adequate lines of credit from foreign banks or at any rate can only do so at a much higher cost than their competitors.

2.2. Intraregional Transfers

The other important infrastructure component for bank competition is the one governing intraregional transfers. Despite the region's high degree of integration in its productive structure, migratory flows and financial system, there are no efficient mechanisms for making payments between one country and another. Intraregional payments have to be carried out through a costly chain of correspondent banks that increase transaction costs and raise portfolio management costs for both local clients and the banks themselves.

How does the system work at present? A bank customer in Costa Rica wishes to transfer a sum of money to a customer of another bank in El Salvador – a typical case involving imports of Salvadoran goods, but which also applies to a remittance from a Salvadoran immigrant. The Costa Rican bank holds an account in at least one correspondent bank in Miami or New York, as does the Salvadoran bank. To make a transfer from the Costa Rican bank to the Salvadoran bank, the Costa Rican bank requests its correspondent bank in New York or Miami, through a swift message, to debit its account and transfer the sum in question to the correspondent bank of the Salvadoran bank, which, once notified, will credit the amount to the receiving customer's account. Commissions have to be paid which in the chain as a whole can amount to more than 5% for the client. Obviously the accounts in the correspondent banks have to have funds, and those resources are generally managed by the correspondent bank. This transfer mechanism makes all regional transfers more costly and imposes time restrictions on the operations: although local banks close at 6 p.m., interregional transfers must be made before 1 p.m. if they are to be executed on the same day, otherwise the funds will not be credited to the receiving account until the following day.

The regionalization of banking groups has made it possible to overcome such limitations, for they have developed compensation systems between their various banks which enable a customer of one of the group's banks in Costa Rica to make transfers to a customer of another of the group's banks in El Salvador, with more flexible hours and relatively lower rates – or in any case, rates that are not comparable to the chain of correspondent banks but are left entirely to the regional banking group. This regional transfer mechanism through compensation between the group's banks places regional groups in a strong position to attract clients engaged in intraregional business, which is part of the motivation behind the regionalization process that has been taking place in recent years. There are less efficient mechanisms than those available with present technology. If the countries were interested in greater regional integration and competitiveness of their banking systems, they would have to update such technologies.12

3. Degree of Connectivity

The evolution of communications and banks' technological development, which have paved the way for e-banking, has radically transformed the shape of financial markets and the different groups' strategies. Five years ago, in its report on consolidation in the financial sector, the Group of Ten stated that: "The continued evolution of electronic finance could expand greatly, or even eliminate, existing geographical limits and lower entry barriers, thereby altering the potential effects of consolidation" (Group of Ten, 2001, p.10). Traditional concentration indicators no longer reflect the degree of competition to which local banks are subjected; after having been protected by laws and different barriers to entry, these banks are rather suddenly facing an intensive competitive dynamics.

Transaction costs limiting local banks' exposure to external competition continue to exist, particularly in certain market segments, but a growing segment of local banks' activities is subject to strong competition from foreign banks and in general from foreign intermediaries.

On the one hand, local investors have the choice of opening accounts abroad and having their capital managed there through stockbrokerage firms in Miami or New York. But the majority of investors who invest in foreign markets do so through the local banks with stock market branches abroad, with which they can share management fees. This loyalty to local banks is explained by transaction costs and the individualized attention provided by local brokers. In fact, to operate directly through a foreign stockbrokerage firm, there are relatively high minimum sums required for opening accounts, which small-scale investors have difficulties to meet. Furthermore, these agencies usually give individualized attention to large investors only and, therefore, the medium-size investors and even some of the

12 The IMF (2006, p. 3) has repeatedly pointed out that the "efficiency gains that would be reached through the adoption of regionally integrated payment frameworks and stock market settlement".

large ones prefer to rely on the local banks. Hence, the competition from foreign intermediaries is restricted by these barriers and is an obstacle for the direct relationship with local investors.

On the other hand, sufficiently large local companies have the option of issuing American Depositary Receipt (ADR)s on Euro markets, as has been done by some of the region's breweries and cement industries, among others. But that particular segment includes a very small number of companies, although the amount of their financing needs is very high. Those companies, as well as a segment of smaller and medium-sized companies, receive offers of credit from foreign banks, which visit them regularly through their local offices when this is permitted, or through salesmen who make periodic visits, just as foreign brokers visit investors. The cost of dollar credits depends partly on the country's risk, partly on the company's soundness and partly on how far its links to the foreign bank date back, but in general it tends to be lower than the cost of credits offered by local banks.

Recent research shows the perception on the degree of openness and development of the financial systems of various Latin American countries, among them those of Central America, Mexico and Panama. Table 4.2 shows this perception for a set of indicators on the degree of exposure to competition and international markets.

As we can see, the predominant view is that it is very easy for investors to invest in both foreign and local markets, but much more difficult for companies to obtain financing in foreign and local markets. This confirms that in order to secure the financing they need, local companies continue to be relatively dependent on

Table 4.2 Selected countries: perception on openness (Arnoldo Camacho, Finance Alternatives in Latin America. Recent Developments and Future Outlook, Sumaq Summit, 4 de mayo de 2004)

a Local citizens who wish to invest on stocks and bonus and open banking accounts in other country (1 = forbidden to do it, 7 = free to do it).

b Foreigner investors (1 = forbidden to invest on stocks and bonus in the country, 7 = free to do it).

c Local enterprises may get into debt in bonus foreign markets (1 = practically impossible, 7 possible enough for a good company).

d Capture money emitting actions in the local market (1 = practically impossible, 7 = possible enough for a good company).

e The entry of new banks in the local banking industry (1 = difficult enough and rarely allowed, 7 = easy and subject to reasonable regulations).

domestic banks, a perception reinforced by the view that it is still relatively difficult for new banks to enter the market.

4. Internal Sources of Contestability

Apart from the competitive pressure arising from the globalization of the financial system and exerted by foreign banks and stockbrokerage firms, banks face various local sources of contestability. The most important in the region, which operate especially in Mexico and Costa Rica, are public debt and the stock market, the latter not so much in the financing of enterprises as in deposit-taking from the public through mutual funds, which in turn invest mainly in public debt.

4.1. Public Debt

Public debt is a source of competition for banks in attracting savings from the public. The returns on public debt are generally higher than those offered by the banks' term deposits, and the public in all the countries of the region have been able to purchase public-debt securities at the Finance Ministry, Central Bank and commercial bank windows. This has been an investment alternative for medium and large investors who invest part of their portfolio in local public debt purchased on the spot market (auctions and windows) or on the over-the-counter market through banks and stockbrokerage firms. This involves a market segment for which the banks are unable to compete, for in order to capture it they should expect to see their funding costs increase appreciably and in any case they already have the most captive market for sight deposits (current accounts) and term deposits from small investors who have no access to public debt. Indeed, public debt has not been an investment option for the latter due to the lack of a distribution network for such bonds, which is explained by the cost for the state of issuing physical certificates. The struggle to attract small investors and sight deposits has been opened by the development of jointly administered instruments, particularly mutual funds.

4.2. The Stock Market

The stock market offers two sources of contestability for the banks: through attraction of savings and through company financing. However, they have been unimportant in most of the countries of the region except Mexico and Costa Rica.

With regard to company financing, in no country of the region, not even Mexico, has the local stock market seriously competed with bank credit. In Mexico the capitalization of the stock market has been falling since 1994, and corporate-debt placements represent a low percentage of GDP. In Costa Rica the relative importance of the stock market and corporate acceptance is even lower. In both countries the stock market basically consists of public debt and bank securities. In Panama stock market activity is virtually non-existent and the high rate of stock market capitalization is accounted for purely tax reasons: the stock issued by companies is a mechanism for receiving bank credit and benefiting from more favourable tax treatment.13

As regards deposit-taking, the stock market has played a certain role in the development of the banking system in Mexico and Costa Rica, although that role has been uneven. As we have mentioned, in 1949, Costa Rica established a state monopoly on deposits from the public which for many years prevented financial intermediation by private concerns. This situation, coupled with the existence of a high level of public sector internal debt, favoured the emergence of a stock market which enabled the state to meet its financing needs and allowed private financial capital to compete with the public banks in the colón market (the colón is Costa Rica's currency). Costa Rica established a stock market long before the rest of the Central American countries and promoted two instruments which were to play a key role in the development of the banking system. First, joint investment portfolios administered by stock exchange seats, the design of which was copied from Mexico. Investment in public debt was done through these portfolios and it yielded higher returns than those offered by the public banks. This was a bank instrument, for it formed part of the stock market seat's balance and guaranteed previously agreed returns and ready cash. Second, it permitted the borrowing of public-debt securities at relatively short terms, thus constituting a means for investors to invest at very short terms with relatively high returns. These two instruments and foreign exchange activities as well as the support from the Agency of International Development (AID) led to the development, during the 1970s and 1980s, of private financial capital. Credit institutions and subsequently banks were established, which took deposits from the public by issuing negotiable investment certificates on the stock market, an instrument developed to replace term deposits, which were prohibited to them until the 1990s. Even at present, the banks used both joint portfolios and the borrowing of securities to manage their liquidity in the absence of an interbank market.

Thus, in Costa Rica the stock market constituted a means for the development of private financial capital and private banks in the face of the public banks' monopoly on deposits. This made it possible for the market's infrastructure, regulation and supervision to make more headway than in the rest of Central America and, at a certain point, to become a source of competition for the banking system, especially with the development of mutual funds. This instrument was initially developed by non-bank stock market seats, and its rapid development, particularly after the legislation passed in 1998, gradually obliged all the financial groups to create mutual funds and propose them to their customers. This instrument replaced the former joint portfolios – which were prohibited – and has been a source of competition in the

13 Tax treatment for companies listed on the stock exchange and their dividends is highly beneficial. Hence a considerable portion of bank credit has taken the form of purchase of share issues which the banks themselves structure and which never pass through the secondary market.

Costa Rican financial system, and its growth over the past five years, as such, reached almost 40% of the banks' liabilities. Some of them, though, in the absence of an interbank market use these as an instrument to handle their own liquidity.

In Mexico the importance of the stock market in the financial system increased following the 1982 nationalization of the banks. As a result of the latter, a growing percentage of bank deposits was earmarked for financing the fiscal deficit, which displaced the private sector from bank credit, and restrictions were imposed on lending and borrowing rates and other quantitative and qualitative limits on credit. Companies' need for funds encouraged the creation of a parallel banking system. In this regard, stockbrokerage firms were the main institutions competing with the nationalized banks, both in deposit-taking and in channelling savings. Their development at that time has allowed them to position themselves with a series of jointly administered instruments which are invested mainly in public debt and to a lesser extent in bank instruments. During the second half of the 1990s and in more recent years, the legal framework for mutual funds has been improved and that instrument has been gaining ground over the banks in deposit-taking.

In the rest of Central America the development of the stock market has been incipient, due largely to the fact that the financial system is dominated by the banks. In El Salvador there is a type of joint portfolios, similar to Mexican and Costa Rican portfolios, which enables stockbrokerage firms to attract sight deposits which they invest in public debt, and in Honduras the stockbrokerage firms issued securities backed by public debt, in both cases earning from the spread. In Panama and Guatemala mutual funds exist, in the latter country with deficient regulation, but in neither of the two countries have they succeeded in attracting significant amounts.

III. Competitive Strategies of the Banks in the Region

A series of common actions by the banks could be taken to mean strategies aimed at strengthening their competitive situation. The first strategy was the reorganization of, and modernization of, the banks to be able to meet the Basel standards and the increasing exposure to competition from foreign banks and stockbrokerage firms. The second strategy has been an active policy of mergers and takeovers as a means to expand operations, attract more clients, enter new fields of operation and eliminate competitors. The outcome of this strategy has been a significant reduction in the number of operators and an increase in the five main operators' share of total assets in each country. The third strategy has been regionalization which has taken place together with the mergers and acquisitions. With few exceptions, the main operators in each country are setting up operations in the other countries, thus accompanying the regionalization of Central American industry and becoming regional operators. The fourth strategy has been the establishment of offshore operations which makes it possible to operate at lower costs for clients and bank stock holders. The fifth strategy has been the combination of specialization and diversification of the products and services offered by the banks.

1. Efficient Organization and Modernization of the Banks

There has been a generalized improvement in the banks' operative efficiency. However, the intensity of this improvement varies substantially. El Salvador during the 1990s is a case in point: as we can see in Table 4.3, in 1997, the administrative costs/total assets ratio stood at an average of 3.42%, far below the other countries; since then and until 2005 that indicator continued to improve until it reached 2.81%, which was surpassed only by Panama. This country had an administrative costs/total assets ratio of 1.10%, a long way from the other countries being studied, in spite of that indicator's deterioration since 2000, when it reached 0.78%. Efficiency levels in Panama are associated with the existence of the financial district, where large part of the operations are barely registered and do not involve costs as high as local operations.

Nicaragua's experience contrasts sharply with that of El Salvador. These two countries' banks are the most active as regional operators, and they would therefore be expected to share similar efficiency levels. This, however, is not the case: Nicaragua features among the countries with the most inefficient banking systems, far below El Salvador, and without registering any significant improvement during the period from 1997 to 2005.

Mexico's case is paradoxical. The majority of the country's largest banks were taken over several years ago by some of the main bank operators at international level. It was to be expected that the high efficiency levels that characterize those operators in their countries of origin and in the other countries where they have operations would be transmitted to their activities in Mexico. This, however, has not been the case. As we see in Table 4.3, operative efficiency measured by the administrative costs/total assets ratio reaches 4.7% in Mexico, just below Costa Rica and Honduras. This situation can be explained by lax regulations in this matter and the existence of a strong trade union movement which has been successful in defending its achievements. The high fluctuation shown by the indicator in this case is also noteworthy.

Table 4.3 Central America and Mexico: operative efficiency of the banks, 1997–2005 (administrative costs/total assets average) (Central American Monetary Council at www.secmca.org/Estadisticas_Indicadores_Ban.htm and Mexico's National Banking and Securities Commission, at www.cnbv.gob.mx)

Note: Figures are for December each year from 1997 to 2004. For 2005, the figures are for June, except for Panama, which are for April.

In Costa Rica, the low operative efficiency evidenced by Table 4.3 is associated with the public nature of its main banks, which are subject to regulations inherent to the public administration with regard to procurement and personnel management, not to mention that they often have to produce results reflecting social rather than profit-making criteria, such as the extent of the branch network. Finally, the case of Honduras, last in the ranking, reflects the country's general backward state and the lower level of development of its banking system. In this context it is interesting to look at Table 4.4, which contrasts various efficiency indicators for Honduran-owned banks with foreign banks operating in the country. The table clearly shows that these have a better performance than the domestic banks in all the indicators.

The table above shows that the average CAR is approximately 4% points higher in the foreign banks than in the domestic banks; the bank with the highest CAR (38.65) is a foreign-owned bank, whereas the one with the lowest CAR (10.36) is a domestic bank. The foreign banks, on their part, generate greater average financial returns (FR) and in turn register a lower cost of liabilities (CL), and consequently their financial intermediation mark-up (9.69%) is 1.5 times higher than in the domestic banks (6.12%). This is related to two facts: higher administrative costs, but higher profitability of foreign banks. In fact, the relationship between administrative costs and average productive assets of foreign banks is 13% higher than the domestic banks, probably because of staff expenses associated with the hiring of international personnel. The returns on productive assets (RPA) of foreign banks is double that of the domestic banks and the returns on assets (RA) is approximately 14% higher (Tábora, 2007).

2. Mergers and Takeovers

As a result of the implementation of the Basel standards and the increased exposure to international competition stemming from the globalization of the financial system, the banking industry has experienced a major process of concentration.

Table 4.4 Honduras: efficiency differences between foreign and domestic banks, September 2004 (%) (Prepared by Tábora, 2007)

a Capital adequacy ratio.

b Financial returns.

c Cost of liabilities.

d Intermediation margin.

e Returns on productive assets.

f Returns on assets.

In addition, the deficiencies in regulation and supervision suffered by various countries throughout the 1990s created the right conditions for a series of crises which stepped up the concentration process. As can be seen in Table 4.6, the five main banks' share of total assets is very high. Nevertheless, the consequence of the relatively similar sizes of the main operators was that the Herfindahl-Hirschman Index (HHI), which measures all the individual firms' share of the industry,14 stood at around 2,000 in the majority of cases.

In Mexico, as of 1990 the privatization of the banks led to a period of rapid growth in the activity and in the number of banks. In 1991, there were 12 banks, rising to 22 in 1992, 27 in 1994 and 33 in 1995. The 1995 crisis reversed the process: by 1996 the number of banks was 31, despite government rescue efforts, and continued to fall over the following years to only 20 in 1999. During that period a mass entry of foreign capital began, and the mergers and takeovers coincided with the country's main banks being taken over by foreign institutions. Thus, the Spanish bank, BBVA, which started operating in 1994, absorbed three existing banks between 1995 and 1999: Oriente, Probursa and Cremi; Santander, also a Spanish bank, began operations in 1995 and in 1997 absorbed Banco Mexicano; whereas Citibank, which was already operating in 1991, took over Confia. Meanwhile, in 1999, another three banks, all of them among the country's largest financial concerns, were capitalized by foreign banks: BANCOMER, capitalized by the Bank of Montreal; SERFIN, capitalized by the Hong Kong & Shanghai Bank; and BITAL, capitalized by Banco Central Hispano and Banco Central Portugués. Likewise, both BANCOMER and BITAL absorbed other entities: the former, PROMEX and Unión, and the later, Atlántico, Interestatal and Sureste.15 During the period 2000–2003, the total number of banks fell from 20 to 18, which involved large-scale bank mergers and changes in control from Mexican owners to foreign ownership: BBVA, recently merged in Spain, absorbed BANCOMER, and Citibank acquired Banco Nacional de México (BANAMEX), Mexico's largest bank16; HSBC took over BITAL; Scotia acquired Inverlat and Santander absorbed SERFIN. These developments left all the major Mexican banks in the hands of foreign capital. (see Avalos and Hernández, this volume).

Central America also underwent a process of bank consolidation over the same period, although the different cases vary among themselves. As we can see in Table 4.5, taking all the banks into account, including affiliates of foreign entities, the most

14 Defined as the sum of the squares of the relative shares of all the firms in the entire industry. The result falls between 0 and 10,000. Zero indicating perfect competition and 10,000, absolute monopoly. The Department of Justice considers a market with HHI lower than 1,000 as competitive; with a value between 1,000 and 1,800 as moderately concentrated and 1,800 or greater as highly concentrated.

15 Throughout the period analysed (1990–2003), foreign banks' affiliates were not taken into account.

16 The operation involved resources in excess of US$12 billion (see Avalos and Hernández, this volume).

Table 4.5 Central America and Mexico: evolution of the number of banks (Central American Monetary Council at www.secmca.org/Estadisticas_Indicadores_Ban.htm and Mexico's National Banking and Securities Commission, at www.cnbv.gob.mx/default.asp?com_id = 0)

Note: From 1998 to 2004, the figures refer to banks in existence in December each year; for 2005, to those in existence in June.

intense process of reduction took place in Nicaragua, where the number of banks dropped from 12 to 6. As in Mexico's case, the concentration process in Nicaragua took place after a period in which the banking system experienced major growth: as of 1990, following the end of the Sandinista regime, ten private banks were created while at the same time the importance of the state banks declined. The banking sector became consolidated with the entry of Nicaraguan-owned private banks which in 1999 concentrated over 90% of the system's deposits, with a relatively well-balanced distribution of assets and liabilities. However, in the year 2000, the Nicaraguan financial system was shaken by a series of bankruptcies that threatened the country's financial stability. The crisis broke suddenly with the takeover by the Superintendence of the country's largest bank, Banco Intercontinental (Interbank), which at the time held 14% of the system's total assets.17 The Interbank intervention had serious consequences as regards account holders' confidence in the stability of the banking system. Between June and December 2000, an outflow equivalent to 8.2% of the public's deposits was registered, reducing the total during the same period from US$1.539 billion to US$1.414 billion. Thus, other banks in the system which were already facing difficulties were affected by this withdrawal of deposits; the loss of credibility combined with portfolio irregularities led to the collapse of another four banks18 (Ansorena, 2007).

In view of the size of the institutions affected, the authorities decided to guarantee the totality of deposits and the majority of liabilities. The assets and liabilities were transferred to other local banks by auction, whereas the Central Bank of

17 What appeared to be a well-managed portfolio turned out to be a situation in which 80% of loans, mostly unrecoverable, had been made to individuals or companies linked to an agro-industrial economic group (Ansorena, 2007).

18 The four institutions taken over by the Superintendence during that period represented in June, 2000, approximately 40% of the system's assets and 39% of its deposits. The liquidity and solvency problems arose as a result of the rapid growth of risk assets, their concentration and the proliferation of related loans, in addition to external factors.

Nicaragua (BCN) covered the difference between liabilities and assets by issuing medium-term internal-debt securities (CENIS) equivalent to almost 16% of GDP. This decision led to increased concentration in the sector: as a result of the auctions, Banco de la Producción (BANPRO) absorbed three banks, Pribanco, Banco Intercontinental and Banco Nicaragüense (BANIC), while Banco de Finanzas (BDF) absorbed Banco del Café and Banco de Credito Centroamericano (BANCENTRO) took over Banco Mercantil. As can be seen in Table 4.6, this process increased the five largest banks' share of total assets to 95.71% in 2001, reaching 99.57% in 2005. This situation was reflected in the evolution of the HHI from 0.09 in 1999 (when there was greater distribution among the nine largest banks) to 1,900 in 2004, above the critical level of 1,800 (Ansorena, 2007).

After Nicaragua, El Salvador was the country that suffered the greatest reduction as a percentage of total national banks during the period 1998–2005. It is also the country, after Nicaragua, showing the greatest asset concentration: in 2005, the five largest banks held 91% of assets (see Table 4.6). This is expressed by the pronounced growth of the HHI for total assets (see Table 4.7). In this case the mergers

Table 4.6 Central America and Mexico: five largest banks' share of overall assets, 1997–2005 (%)a (Central American Monetary Council at www.secmca.org/Estadisticas_Indicadores_Ban.htm and Mexico's National Banking and Securities Commission, at www.cnbv.gob.mx/default.asp?com_id = 0)

a Equivalent to the five main banks' share of the total banking sector in each country.

Table 4.7 Central America: Hirschman-Herfindahl index total assets, 1996–2005 (Based on official sources from Ávalos and Hernández in this volume, Balsells, 2007, Herrera, 2007, Tábora, 2007 and Yong, 2007)

a Equivalent to the HHI of total credits.

b Figures correspond to March.

and takeovers process did not stem from a financial crisis, but from strategies aimed at strengthening the banks' positioning in the domestic market and attaining scales of activity to enable them to operate at the Central American level. The first merger took place in 1997 and involved an international bank from outside the region: in October that year The Bank of Nova Scotia (Scotiabank) acquired 53% of the shares of Ahorros Metropolitanos, making it the first international bank to hold a majority share of a Salvadoran bank. Nova Scotia increased its shareholding in Ahorromet to 98.3% in December 2000, when it changed its name to Scotiabank El Salvador. In August 1998, two small banks, PROMERICA and Banco Corporativo (Bancorp),19 merged. In July 1999, BANCOMER took over Banco Atlacatl. In May 2000, the merger began between Banco Agrícola Comercial, the country's largest bank (with a 20.27% share of total credits, 24.55% of the deposits market and 23.04% of the system's overall assets) and Banco Desarrollo, which in turn was in fifth place in the system. In July 2000, the merger between Banco Salvadoreño and Banco de Construcción y Ahorro (BANCASA) took place.

In Guatemala's case, the number of banks dropped from 34 in 1998 to 25 in 2005. The reduction in the number of banks was associated with the 1998 crisis20 and with corporate strategies aimed at strengthening the institutions' positioning in the domestic market. In 1999, a period of mergers began which continues to this day. That year Banco del Café absorbed Multibanco, thus rising to second place by size of assets; in 2000, Banco Reformador took over Banco de la Construcción, passing from eighth to fourth place in the same ranking; in September, authorization was granted for the merger between Banco del Agro and Banco Agrícola Mercantil, forming the new Banco Agromercantil de Guatemala, second place in assets that year; in March 2001, Banco Granai & Townson and Banco Continental merged, thus creating Banco G & T Continental, which became the country's principal bank; in November, 2002, Banco CHN took over Banco del Ejército, as well as another small bank, Banco Del Nor-Oriente (BANORO), the following year; finally, in April, 2004, Lloyds TSB Bank, Guatemala Branch, made over the totality of its credit assets in favour of Banco Cuscatlán de Guatemala, as a result of which Banco Cuscatlán rose from eleventh to eighth place in the market upon increasing its assets by 50% (Balsells, 2007).

The mergers process in Guatemala began in a context of low concentration levels. In fact, in 1998 the five largest banks' share of overall assets was 38.84%. Despite the major increase in the five largest banks' share of total assets in 2005 (62.15%), Guatemala continues to register the lowest concentration levels, the only exception being Panama. Analysed on the basis of the HHI with respect to total credits, the

19 Banco Promerica established its operating strategy through alliances with important chains of supermarkets and shoe stores, Pricesmart and Payless Shoe Source in various Central American countries. This bank's equity is 99.99% in the hands of Salvadoran shareholders according to data in the Boletín Estadístico, December 31, 2004 (Herrera, 2007).

20 In 2001, three private entities were intervened due to causes related to the 1998 crisis. Those banks were Banco Empresarial (February), and Banco Metropolitano and Banco Promotor, both in March that year (Balsells, 2007)

index has never risen above 881, which denotes low levels of concentration. The other characteristic aspect is that apart from the last case involvi