POLIC;Y RESEARCH WORKING PAPER 23 18 RegLilat ory Reform, Regulatory reform can spur innovations in infrastructure Comapetition, and service;, generating new Inn ovaticon downsi:ream activities and magnifying the economywide benefits of reform. The A Case Study of the M exican national competition agency Road Frelght -ndustry can help greatly in laying the groundwork for reform by makin(g a compelling case For Mark A. Dutz the reform's expected Aydin Hlayri benefis. Pablo Ibarra The World Bank Develcpment Research Group Public Econonmics April 2000 POLICY RESEARCH WORKING PAPER 23 18 Summary findings Discussions of competition and regulatory reform company, the benefits of reform included a IO percent typically focus on price and quantity effects. But improvement in operating margin. improving certain infrastructure services can also Successful reform requires careful planning and stimulate entry and competitioni in user industries execution and political support at high levels. Regulatory downstream, allowing new firms to enter, incumbent reform also profoundly changes the sectoral institution users to offer new products, and rivalry to intensify. formerly responsible for the regulation. Enough Dutz, Hayri, and Ibarra present a case study of how resources should be provided to help organizations in the innovations in road freight services affect selected reformed industry make the transition to the post-reform downstream users of those services after regulatory environment - helping witlh such tasks as defining the reform. After a period of rigid regulation and hieavy organization's new role and facilitating the redeployment government interference, Mexico in 1989 developed a of staff. new policy framework for road transport, with free entry The national competition agency can help greatly in and market-based price setting. The result: faster, more laying the groundwork for reform by making a reliable trucking has allowed user companies to offer compelling case for the reform's expected benefits. After new, previously unavailable products and to reach new reform, the competition agency should also help with areas with existing products. Cheaper, more customer- enforcement, to ensure that the cozy, cartel-like behavior responsive trucking services have allowed logistical stimulated by tight entry restrictions does not persist. In innovations in user firms, and some user firms have Mexico, three strong interventions were requi:red to decided not to keep their own fleets of trucks but to discipline attempted anticompetitive practices in the outsource trucking services on the open market, thereby trucking industry in the years following reform. converting fixed costs to variable costs. For one fertilizer This paper is a product of Public Economics, Development Research Group. The study was funded by the Bank's Research Support Budget under the research project "Competition and Barriers to Entrepreneurship" (RPO 682-57). Copies of this paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Hedy Sladovich, room MC2-609, telephone 202-473-7698, fax 202-522-1154, email address hsladovich@worldbank.org. Policy Research Working Papers are also posted on the Web atrwv-wwv.worldbank.org/research/workingpapers. The authors may be contacted at mdutz@oworldbank.org, ahayri@dttus.com, or ibarrarodriguez_pablo@jpmorgan.com. April 2000. (48 pages) The Policy Research Working Paper Series dissenzinates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even ifthe presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Banzk, its Executive Directors, or the countries they represent. Produced by the Policy Research Dissemination Center Regulatory Reform, Competitioin, and Innolvation A Case Stucly of the Mexican Road Freight Industry Mark ]Dutz, Aydin Hayri and Pablo l[barra The invaluable field work assistance of Guillermo Abdel Musik (as survey team leader) and Sergio Garza, both at the Centro de Estudios Estralegicos at ITAM, MeXico City is gratefully acknowledged. The paper benefited fromr helpful suggestions by Cesar Cordova-Novion, Scott Jacobs and Benedicte Larre. The project was prepared under funding from the World Bank research preparation gramt 'Competition and barriers to entrepreneurship' (RPO-1682-57). This research reflects the work of the authors and does not reflect the views and policies of the World Bank nor the countries it represents. Correspondence: mdutz(@worldbank.org, ahayri@dttus.com, pibarra@princetoi.iedu. Outline Executive Summary 1.0 Introduction 2.0 Regulatory Struct.ures and Their Reform 2.1 The Political Economy of Regulatory Reform 2.2 Regulations and Related Policy Instruments 2.2.1 The Preregulatory Reform Environment Market Access Market Conduct Prices 2.2.2 Regulatory Reform Market Access Market Conduct Prices 3.0 Market Stiructuire and Entry 3.1 Evolution of Industry Structure 3.2 Industry Conduct and Intensity Of Competition 4.0 Performance 4.1 Impact of Reform on Performance in the Road Freight ][ndustry 4.1.1 Output and Prices 4.1.2 Innovations and Productivity 4.1.3 Profits, Wages and Employment 4.2 Imlpact of Reform on Performance in Downstream Industries 4.2.1 Costs 4.2.2 Innovations and New Products 4.2.3 Additional Ecornomywide Impacts 5.0 Policy Options For Reform 5.1 Successes and Failures of Reform 5.2 Potential Benefits and Costs of Further Regulatory Reforms Appendixes Appendix 1. The Road Freight Industry in Mexico Appendix 2. Structured Survey Instrument Example: Fertilizer Company Appendix 3. Foour Case Studies Figures Figure 1. Road Freight Hau:lage Units of the Mexican Public Federal System, 1980-1996 Figure 2. Road Freight E,nterprises According to Number of Haulage Units, 1996 Figure 3. Quality-Price Tra(le Off in the Road Freight Industry Figure 4. Age of Equipment (Motor Units) 1990-97, and Projections, 1998-2010 Executive Summary Typically, discussions of the benefits of competition and regulatory reform focus on price and quantity effects in the market under consideration. However, improvements in certain infrastructure services also can stimulate entry and competition in do-wnstream user industries, allowing new firms to enter, incumbent users to offer new products, and rivalry to intensify. To the extent that refor m spurs innovations in infrastructure services, and these innovations in turn generate substantial new downstream activities, the econoniywide benefits of regulatory reform are likely to be substantially greater. In this case study, we quantify the positive benefits achievable from innovations in road freight services on selected downstream firms that are intensive users, of such services, following regulatory reform. From a particularly extreme degree of rigid regulatien with a high degree of government interference, Mexico put in place beginning in 1989 a new -policy framework for the road transport industry based on free entry and market-based price setting. Besides expectecd gains from reductions in trucking prices, there have been a number of additional sources of downstream gains. Importantly, faster and more reliable trucking has allowed user companies to offer new goods, both introducing previously unavailable products as well as making it possible for existing products to reach new areas. Cheaper and more customer-responsive trucking services have also led to a number of other logistics-related innovations within user f.-irms. Finally, some user firms, instead of co:ntinuing to invest and maintain a private own-account fleet of trucks, have outsourced and purchased the services on the now more efi'icient open market. Although such costs are not largely sunk, tapping capital markets may not be an option for prospective entrants lacking reputaltion. An important additional benefit of a more competitive trucking service market, therefore, is to allow firms to outsource their transport requirements, in effect converting fixed into variable costs. The total up-side benefits are sizeable: the improvement in operating margin d:irectly attributable to road freight innovations for a representative fertilizer company is a surprisingly large 10 percent. An important political economy lesson is that successfiul reform requires careful planning, execution and high-level political support. Regulatory reform also causes profound changes in the sectoral institution formerly responsible for the regulation. Any regulatory reform effort should assign sufficient resources to assist such organizaticn in making the transition to postregulatory reformi conditions, assisting with such tasks as defining the organization's new role and facilitating redeployment of staff. Finally, the national competition agency can play a critical role in helping lay the groundwork for reform, by nmaking as compelling a case as possible for the expected benefits of reform. In addition, the competition agency rnust play a careful post-reform enforcement role to ensure that cozy cartel-like behavior stimulated by tight entry restrictions does not persist. At least three separate strong interventions were required to discipline attempted anticompetitive practices by the trucking industry in the years following the initial regulatory reforms. 1. Introduction The transport of freighit is an input to the production of capital, intermediate, and consumer goods and services. As can be verified by input-output tables, it is almost impossible to think of a line of business that does not use transport. As such, any inefficiencies in transportation services as well as any monopolistic structures in its markets constitule an additional cost and at times a bottleneck for all other sectors of the economy. WVhen adding unnecessary costs, inefficient transport services are equivalent to additional import or export tariffs. When constituting a bottleneck due to lack of capacity to move more freight in a more tirmely manner, they become equivalent to quantitative restrictions. Importa:ntly, in addition to being seen as a means to reduce costs, more efficient transport and logistics systems are increasingly regarded as a means of boosting revenues and a source of competitive advantagre. With the increasing emphasis of businesses on revenue-enhancement opportunities from logistics management and outsourcing, inefficiencies in freight transport can cripple a country's efforts to enhance its overall competitiveness. IBox 1] In the Mexican transport sector, road freight transport oni federal highways prior to 1989 was subject to a rigid regulation with a high degree of interference by government. Important government-imposed barriers to competition included entry restrictions to operate on federal highways, discretionary allocations of freight among truckers, and strong restrictions on moving cargo outside the established transport corridors. Of'ficial tarii'fs applied to all cargo and a semipublic company held a monopoly in handling containers. Regulations did not allow companies to cha:rge higher rates for better service and hence no incentive to offer better services. Neither did they allow them to compete with one another by offering lower rates. As a result, the trucking indu,stry was characterised by a limitec. number of firms operating with minimal competilion. Moreover, to maintain this highly ineifficient and archaic system, the government employed a sizeable bureaucracy. In 1989, the govern-ment formulated a new policy framework to deregulate the road transport industry. The government's vision was to create a competitive trucking industry with free exit and entry and market-based pricing, without intrusive government regulation. The government focused its deregulation efforts in transportation and telecommunications with the belief that they accounted for major supply bottlenecks and market im.perfections. Their reform was deemed essential to achieve more rapid and sustainable growth. This study assesses the impact of regulatory reform in the Mexican road freight industry by focusing on selected dlownstream users. Following an overview of the political economy of regulatory reform, the evolution of regulatory structures is examined to provide a context for the subsequent inmpact- assessment. The market impact of regulatory reform is assessed first by examining the evolution of market structure and conduct, and then by quantifying changes in performance. The: major theme in the assessment of perfonrmance is the positive impact of removing bottlenecks for competition. In particular, we emphasise the large magnitude of downstream benefits possible from innovation and new products that accompany rmore intense rivalry in the provision of this essential upstream business s -rvice. The study concludes by highlighting remaining areas where further regulatory refonn may yield additional economywide benefits. 1 Box 1: Supply Chain Logistics - Moving Up the Corporate and Policy Agenda Transportation and the somewhat broader concept of logistics (including warehousing, inventory and administration in addition to physical transport) have long been seen as minor cost items. Businesses did not believe that they could build competitive advantage based on logistics and hence ignored it. Similarly economic analysis recognised only the direct benefits of lower transportation costs. That was before the supply chain revolution in the industrialised world (supply chain management refers to the integration of the flow of gooc[s, information and finance to move products from suppliers to manufacturers to retailers as efficiently and inexpensively as possible). First came the innovation ofjust-in-time inventory systems in 1980s. The idea was simple: the bigger a firm's inventories the larger its operating capital requirements. With more cost-focused management, such as ordering inputs just when needed, companies leamed to slim down inventories. Importantly, they also realised that the infrastructure for just-in-time input deliveries could be useful for quickly changingy designs and customising products. More cost-focused inventory management and flexible production technologies went hand in hand, leading companies to realise that supply chain management is not simply a means to redulce costs but a means to gain competitive advantage over rivals. Companies introducing overnight deliveries to their retailers and next-day service for their customers began to win substantial market shares from established competitors. The dynamic aspects of supply chain management for a manufacturer require feedback from retailers on daily sales mix and volume. The manufacturer responds to the feedback by changing design, sourcing and production volumes. Flexible manufacturing became the darling of business academics towards the end of 1 980s.' A key component of this process was the development of an efficient communications interface with suppliers, cutting back on transaction costs. On-line computer systems allowed companies to communicate with their suppliers as easily as they do intemally. Efficient communications, however, lead to maximum gains only if goods flow as quickly as information. Logistics and communications are highly complementary factors especially for businesses that manufacture or distribute and market consumer goods. Top European industrial and service sector companies have reduced logistics costs to 7.5 percent of revenues in 1998 from 14.3 percent in 1987, and cycle times to 12 days from 27.2 As companies have leamed to establish flexible links to their suppliers and customers, they grew more confident about aggressive outsourcing. It is now a mantra of management that except for a few core functi cins everything else should be outsourced.3 Remarkably, about 70 percent of all U.S. Fortune 500 companies outsource some or all of their logistics activities.4 Outsourcing has an important feedback effect on logistics. Without outsourcing producers must either establish their own in-house distribution networks (costly to set up and difficult to manage) or rely on distributors or end-users to accept delivery at factories. Using third-party distributors or public warehousing and transportation facilities, even a small producer can have access to an extensive distribution network and pay only for usage. This brings down entry barriers created by large capital requirements, especially significant for countries with poorly performing capital markets and for prospective entrants lacking reputation. Despite improvements in financial markets, capital requirements remain as the only entry barrier that has a robust, positive correlation with supernormal profits. Therefore, the conversion of fixed costs into variable costs should spur competition in many industries. In the new world of flexible manufacturing and outsourcing, linkages are becoming more important. Information and goods now flow faster and more accurately. As companies can exploit market opportunities without significant fixed costs, they compete more vigorously. Only companies that are able to communicate and interact efficiently and effectively with others stand a chance of expanding their business. They too have tCl share these benefits with consumers in the form of better service and lower prices. The backbone of this system is third- party providers of efficient logistics and communication services. Using their services, any firm, no matter how small and where located, can reach world markets. In order to facilitate the benefits of the logistics revolution in their countries, policy makers should understand how third-party suppliers of logistics services emerge. This study is a first step in that direction. By taking a deeper look into the logistics operations of downstream user companies in Mexico that could have taken advantage of the recent deregulation of the road freight industry, we identify potential benefits as well as opportunities for improvement. i See Sable and Piori (1990) on the emergence and relevance of flexible manufacturing. 2 See AT Keamey (1998). 3 See Porter (1997). 4Based on an October 1998 survey for the US Council of Logistics Management, as reported by E. Alden, 'Complexity prompts calls for third party', in Financial Times, Supply Chain Logistics, FT Survey, 1998. 2 A novel part of this study is that it benefits from an in-depth survey of 30 firms, 15 trucking companies and 15 intensive users of trucking services in Mexico's heartland undertaken in midl-i998.5 The survey of shippers focused or, three types of industries representing different cormmon users in Mexico: agroindustrial, electronics and auto-parts firms, as well as general import distributors. The most detailed survey instrument combines technical data and subjective assessments of respondents in a consistent framework, thereby generating a set of numbers measuring the value to individual downstream firms of key upstream innovations associated with regulatory reform.6 2. Regulatory Structures and Their Reform 2.1 The political economy of regulatory reform Rationales for regulation. Government regulation of the trucking industry in Mexico extends back to thie late 1940s. However, lack of enforcement of existing regulation since its origin until 1977 was the main characteristic of this period. Most of the trucking companies did not have legal registration and most of the trucks did not have official authorization to circulate. In 1974. more than 7000 trucks did not have the corresponding permit or concession, while only 338 out of 1500 existing companies were legally registered.7 In 1977, a gradual process of increasing reguLlation started with SCT's (the Mexican Ministry of Communication and Transport) Programn of Development for Federal Trucking (Progreama de Desarrollo del Autotrans,porte Federal, i977-1982). The Pirogram [Fostered the revision of the consitutive deeds (escrituras constitutivas) of trucking companies. It also created three different committees: (1) state committees responsible for coordinating actions between SCT and the state authorities, including the construction of freight centers; (2) route committees to emit opinions for the grant of new concessions on specific corridors; and (3) technical committes to review 'broader' aspects- such as economic. industrial, or legal-to be consider for future reformis. Both the r oute and technical committees opened the door for an increasing involvement of trucking leaders in the design of governmental policy. The regulation of roacl freight transport has been normally justified on grounds of safety, reliability of service and stability of rates. In the case of Mexico, the intcreasing regulation was justified on the need to provide a more reliable service and to protect transport infrastructure. 5 Of the carriers, 8 have their headquarters in Mexico City and the othLer 7 in Guadalajara, Jalisco. Of the shippers, 4 :have their headquarters in Mexico City and 11 in Guadalajara. Of the 15 carriers, 6 are small companies with fleets of less than 30 trucks and 8 are larger companies with fleets ranging from 50 to over 400 trucks. The final transpo:rt firm interviewed, a multi-modal freight broker, owns no trucks and is representative of a new segment that operates as a link between Mexican trucking firms and international shipping companies. 6 Our approach does not rely on measuring actual operational results of individual companies like the studies measuring the impact of privatization. Instead, we go through each major operating parameter of the company and construct counterfactuals. Thus we reconstruct each data point rather than simply measuring it. This allows us to achLieve greater reliability with fewer data points. See Appendix 2 for an illustrative example of the most detailed survey instrument. 7 As cited in Islas (1990), p. 117. 3 The apparent rationale was that open competition might: (1) disrupt economic activity through probable high price variability and uncertain service availability especially to less accessible regions of the country, and (2) lead to cost cutting practices that might result in increases in accidents and pollution (due to usage of older trucks, less maintenance, etc.).' However, the changes could also be interpreted as a set of measures that protect the interest of a handful of trucking families, and allowed the latter to implement more strict controls in the market. Before trucking deregulation, about 15 families controlled the whole industry, although it contained several thousand individual truckers.9 Some of the most prominent truckers did not own a truck, but their power came from the control and ownership of freight terminals. As, it happened in 1989, the decision to open the sector to competition unavoidably disrupted some of the most profitable operations. A broad market-based reform context. Trucking deregulation was undertaken within the context of general reform rather than as an isolated action. From 1983 onwards, Mexico began a process of macroeconomic and structural reforms following the debt crisis and the collapse of oil prices. One of the many facets of reforms was to increase the reliance on market forces. International trade was liberalised through elimination of quantitative restrictions and reductions in tariffs as part of Mexico's commitments due to its new GATT membership in 1986. Trade liberalization directly impacted the trucking industry. The collusion of truckers in Mexico was maintained through enforced territorial and cargo distribution of the market. Market distribution responded to transportation flows generated under the import substitution strategies followed for over 35 years. Trade liberalization brought changes to both trade flows and the structure of cargo movements. Exports increased three-fold over 5 years. Cargo increased substantially in the routes connected with international trade (Mexico City- Monterrey-Laredo, Mexico City-Veracruz, and Manzanillo-Guadalajara) but decreased relatively in domestic routes (Mexico City-Guadalajara and Guadalajara-Monterrey).'0 These changes altered the interest of the members of the trucking chamber and truckers in general. Truckers previously locked into unprofitable routes, individual truckers exploited by freight terminals and all downstream users could be expected to be strong supporters of trucking; deregulation. The realignment of interest within the industry played in favour of the Salinas administration's goal of increasing competition and enhancing market contestability through deregulation. Institutional forces for change in trucking. During the period 1988-89, major structural reforms were introduced by the Salinas administration including efforts to expand that miaLrket economy and strengthen the forces of competition. Within this context, the Ministry of Finance (Secretaria de Hacienda y Credito Publico, or SHCP) and the Office of Economic Deregulation within the Ministry of Industry and Trade (Secretaria de Comercio y Fomento Industrial, or SECOFI) argued for the deregulation of trucking. The deregulation of trucking in ' World Bank (1993), p. 8. 9 See Fernandez (1993), p. 105. °0 Idem. 4 the United States in the early 1 980s must have been an impor-tant facto:r because researches generally attached relatively high welfare gains to deregulation. One study, for exarmple, estimated that the gains from deregulation as high as one third of total income generated from trucking." Since the M:exican regulation was even more heavy-handed, the authorities had every reason to expect even bigger welfare gains from deregulation. A1987 report by the Ministry of Commnunications and Transport (Secretaria de C9municacjones y Transporte, SCT) estimated the welfare costs of trucking regulation at 0.'i percent of GNP.12 As part of the larger program of structural reforns to strengthen competition, S]-JCP and SECOFI took the initiative for the deregulation of trucking. The reformrers had to deal with the possibility that the National Trucking Association could have immobili sed the country in their opposition. Although deregulation was expected to benefit a large number of previously disadvantaged truckers, the real power of the trucking leaders had yet to be tested. This threat coupled with the absence of visible strong support from potenitial winners effectively led the reforming bureaucrats to adopt the following gradualist three-stage strategy. This process is yet to work itself out completely, since intra-state trucking still remains heavily regulated and technical regulations were largely ignored at the time. A three-stage approach. In order to avoid interest-group pressures, the draft proposal was prepared by a small group, without open consultations with SCT or other parties. Alter the draft documents were prepared, the first stage of implementation was to solicit support from the affected industry associations, emphasising modernizatioin rather than deregulation without ever openly mentioning strengthening of competition as an explicit objective. In accordance with the recently issued National Development Policy for 19139-94, which outlined lhe general policy framework, the government negotiated an agreement with the national trucking association (Camara Nacional de Autotransporte de Carga, CANACAR) under the terms of which CANACAR. agreed to cooperate with the government in the deregulation and. modernization of the industry. The agreement was signed on July 6, 1989. The agreement stressed the advantages of restructuring and modernising the industry. As a gesture t'o CANACAR, the goveniment offered loans at preferential rates to truckers who wanted to renovate their fleets.'3 The second stage, was the approval by decree on July 7, L989 of new regulations eliminating most e:ntry restrictions. Through this decree, the government effectively abandoned the public service notion that was behind the concept that trucking operations require a concession. The July regulations also allowed 'official' tariffs to be regarded as maxcimum rates. The third stage (deregulation part two) was the issuance of a decree in January 1990 abandoning tariff ceilings thus allowing rates to be freely negotiated between truckers and "' See Moore (1975). 12 As cited in World Bank (1990), p. 1 1. 13 The authorities were reportedly aware that deregulation would lead to a substantial increase in supply, given the excess capacity created by the former regulation (private fleets were prevented from providing service to the broader public, route restrictions created a large number of empty return hauls, ancl freight center controls led to excessively long waiting times). Stimulating further supply through soft klans at that time can therefore nol be explained by economic motives. 5 customers. The principal reason for issuing two decrees separated by six months was SCT' s concern that if all deregulation actions were taken at once, the forces of competition may initially be too weak to prevent sharp price rises. Initial reform by decree. Regulation was implemented by Presidential decree. It was not until late 1993 that the new law was published. The new regulation was only published in late 1994. This delay in publication of laws generated substantial uncertainty among carrier companies as many were not clearly informed of the new rules of the game, or expected an eventual reversal. Apparently, the signing of NAFTA accelerated the effective implementation of these regulatory reforms. The motivation for this phased approach, however, was strategic. It was expected by the reformers that a constitutional reform would take substantial time and meet stiff resistance within Congress. Although the PRI had a majority in Congress, transport association leaders had close connections with representatives in Congress who could have stalled the reform. It was therefore deemed more expedient to initiate reforms through decree (under executive authority) while carefully choosing the wording of texts in order not to contravene existing constitutional pronouncements. As an illustrative example, the prevailing law in 1989 stipulated that the provision of public road freight services requires a concession that must specify the specific route(s) that can be served by the concession holder. The modification introduced by the July 1989 decree eliminated all route restrictions by merely changing the definition of 'route' to 'any route covering the entire national territory'. The period between 1989 and 1993 was seen by the government as a test period during which they could have back-tracked had they faced excessive resistance. In spite of the uncertainty created by ambiguous rules of the game, the approach taken was a pragmatic approach that mininmsised the risk of a reversal. Concessions, permits and the public service provision. To improve transparency and ensure the permanence of the reforms, a radical step was the reform of the law that took pl;ace in 1993. Prior to that date, public road transport services (freight and passenger) had by law the character of a 'public service', which required any nonstate provider to hold a 'concession:' granted by the federal government. Through the concession system, the federal state had full discretion to give and take away authorization to operate vehicles. By basing entry into road transport services on the simple obtaining of a permit, the 1993 law effectively abandoned the public service character of road transport. Through the permit system, the federal state no longer has any discretion regarding authorization but rather must grant it obligatorily as lo]ng as the applicant meets the established requirements. Of greater potential economywide efficiency consequence, the change to nonconcessioned service has automatically transferred, through a constitutional mechanism, the main regulatory powers from the federal government to the respective local governments (states and municipalities) depending on the road the trucker is using.'4 This change has opened the potential for substantial differences in the extent of federal and state-level liberalization, with attendant costs to business arising from lack of harmonization. 14 For a more detailed analysis of these issues, see Sempe (1997) pp. 79-83. 6 Changes in SCT's role. The elimination of entry control and rate regulations greatly reduced the responsibilities and activities of SCT, particularly within the General Directorate of Land Transport and Tariffs. The remaining functions of SC'T with respect to road transport include supervision of road maintenance, monitoring transport and formulating transport policy. SCT also assumed some new functions, such as super,ision of highway safely. However, there was a long delay in reducing the number of personnel. Only by May 1994, did SCT significantly reduce the number personnel in departments previously involved in entry control and rate reg;ulations. SCT assigned the remaining personnel in these departments to others. Reportedly, as the beneficial effects of the new policy have become evident and following the departure of certain SCT staff previously involved in regulatory matters, the commitment of scr to the new policy has become very strong.'5 Deregulation ait the federal level, affecting only interstate trucking. Under the federalist system and the 1993 law, states have autonomy to regulate inlrastate trucking. However, the federal government encouraged state governments to undertake similar deregulation within their own jurisdictions. sc rdevised a plan for harmonising federal trucking regulations with state regulations through a series of bilateral agreements between the federal government and the states. These agreemenl.s were expected to be finalised by March 31, 1991.16 Altflough some states, such as Chiapas, have deregulated intrastate trucking, the harmonization agreements have not been finalised yet. SCT recognises that the progress with bilateral agreements has been much slower than planned. SCT also has doubts regarding the legal authority of the Federal Competition Commission to eliminate intra-state tariffs, schedules and reserved routes. The continued lack of harmonization between federal and state-level regulations, if unaddressed. could lead to a bottom-up backtracking of regulatory reforms, affecting market power as well as environmental and safety objectives -- with the possible rebuilding of cartels at state level, and conflicting stringency requirements regarding environmental, safety and Nveights and dimensions standards. Postreform gradual adjustment-attempts to revive the cartel. The former market intervention policies through segmentation and control of prices and quantities have left a strong imprint on postreform market conduct that has continued to manifest itself. Although regulation by government has been removed, there has been a natural tendency among previously favoured market participants to replace government control by interfirm agreements. The efi-orts of the Mexican Federal Competition Commission have played an important role to oppose anticompetitive practices and complement the deregulation goal of enhancing market contestability. In 1994, the Mexican Federal Competition Commission carried out an ex officio investigation into an alleged absolute monopolistic practice by CANACAR. It involved an agreement among the Chamber's members to fix the prices applicable tc their cargo iransport services. The practiae arose from a reference price guide for negotiations between users and truckers. CANACAR hald written the guide and distributed it among its members. The 15 World Bank (1995), p.33. 16 World Bank (1990), p.16. 7 document specifically established the aim of setting minimum prices for road transport services. This arrangement between competitors was found to constitute a violation of the competition law. As a result CANACAR was fined on its members' behalf. The Chamber also was ordered to withdraw from circulation those copies of the price guide already among its members; to refrain from issuing any other kind of guide with the aim of fixing prices or minimum costs for the services provided by its members; and to desist from establishing pricing policies aimed at creating minimum conditions for the hiring of road transport services. Subsequently, CANACAR consulted the Commission regarding the development of a cost accounting program for its members. This program involved dissemination of a costing method, examples of its use, and training courses or meetings at the national level. To enable members to use the costing methodology more easily, CANACAR prepared a questionnaire for the use of each transport company to identify the elements that make up fixed and variable costs and to estimate the latter. The Commission decided that the use of this costing methodology could lead to greater efficiency in road transport operations - and was therefore not in itself a violation of the competition law. In this context, information on costs could direct transport companies' investments in more efficient units and assist towards a better organization of routes. The Commission did, however, take into consideration the precedent of CANACAR's prior price fixing activity. It therefore warned the Chamber that the example on how to use the methodology would have to be strictly illustrative in nature, avoiding at all times the possibility that the estimates it contained might serve to fix service prices. The Commission ruled that the material used to disseminate the costing method must indicate ithat the use of methodologies and estimates as ways to fix prices between competing road transport companies constitutes an absolute monopolistic practice and, as such, is sanctioned by the competition law. As another example, in 1997 the Federal Competition Commission carried out an ex officio investigation on possible anticompetitive practices in the transport and distribution of diesel fuel. The investigation determined the existence of agreements (1) to divide markets into geographic zones for the purposes of transport and distribution and (2) to limit supply of transport services for fuel oil, in order to strengthen the ability to divide markets. The market division of transport services included two explicit agreements, with the allocation of explicit distribution concessions that acted as entry barriers. In one, transporters divided the transport to consumer enterprises of the product sold directly by Pemex-Refinacion, with the tacit initial acceptance of this parastatal enterprise. The other consisted of the division of transport services of hydrocarbon sold by two distribution companies,'7 with their consent and participation. The Commission declared the practices illegal, imposed dissuasive economic sanctions, and promoted conditions favouring competition, ordering the producer-distributors to assign routs in a transparent manner. Regarding subsidies, some official entities have in the initial years following deregulation still rewarded affiliation to trucking chambers. For example, the National Bank for Public 17 The two companies were Distribuidora de Combustoleo de Cd. Juarez and Distribuidora de Combustoleo. 8 Works and Services (BANOBRAS, Banco Nacional de Obrzs y Servicios Pablicos) reportedly required a letter from CANACAR to ensure the suitability of the trucking company as potential beneficiary for loans at subsidised rates in the early years following deregulation. Our empirical work corroborated that carriers who are not affifiated to trucking chambeirs-often small-owned operators-do not have either the information or the access to any available subsidised loans. Without doubt, the continuing presence of trucking chamrbers has fiacilitated incumbents' control over service supply in some regions. Their contact network has continued to function as an instrument to disitribute higher volume cargo among their members. Neverthe'less, the forceful interventions of the competition agency, together with the development of formal freight forwarders has gradually eroded CANACAR's power, particulazrly in intermodal transport. 2.2 Regulations and relatedpolicy instruments Road transporl: regulation was encompassed in the Law of General Communication Modes (Ley de Vias Generales de Comunicaci6n) and its accompanying implementing regulation.'8 For road transport. this law was replaced on December 22, 1993 by the Law of Federal Roads, Bridges and Trucking (Ley de Caminos, Puentes y Autotrans,orte Federal), and supplemented on November 22, 1994 by the Regulation for Federal Cargo Trucking ('Reglamento para el Autotransporte Federal de Carga).'9 2.2.1 The preregulatory r eform environment Market access. Federal policy strictly controlled entry. Entry regulation differed for public for-hire trucking companies and for private own-account trucking fleels that are owned and operated by firms :in other industries primarily for their own purposes. Main restrictions were on entry to operate on ifederal highways, on cargo movement between established corridors, on cargo assignments, on container handling, and on the handling of cargo at railroad stations and international borders: * Mandatory concessions based on agreement from incumbents. T'o provide public general freight services, truckers were required to obtain a federal trucking concession from SCT. Concessions were specific to class of service, route, number of haulage units that can be used, and technical characteristics. By law, the concessions granted to any individual could not be for the operation of more than five vehicles. Cormpany affiliation for a concession holder was compulsory. If am individual was grarnted such a concession for general freight, he still had to be accepted by an established company and this constituted an impassable barrier to entry. The administrative procedures for determining the need for additional service depended on the opinion of existing service providers. Incumbent firms had both preferential treaktent whenever it was necessary 18 Reglamento al Capitulo de Explotacion de Caminos y de la Ley de Vias Generales de Comunicacion. For a detailed analysis of this regulation, see Davila Capellaja (1994). 19 Dates reflect publication in the official gazette (Diario Oficial de la Federacion,). 9 to increase the number of trucks and the right to object to future increases in installed capacity. These procedures strengthened the position of existing firms and led to the formation of cartels. * Reserved routes. The country was divided into eleven corridors and each trucker needed a concession to operate within a specific corridor. By being route-specific, the concession system controlled installed capacity for each corridor, creating a ceiling to ton-kilometres per route. The corridor system favoured existing spatial concentration since the corridors tended to radiate from Mexico City. Down-time was high and load factors low as the restrictions constrained their options for back-hauling and made it more difficult to match their supply with available demand. For shippers requiring cross-corridor shipments or requiring services beyond the established route capacity, special arrangements with higher tariffs permitting the use of trucks allocated to other routes had to be made. This segmentation was critical to prevent market forces from generating more competition. * All traffic through freight centers. Truckers must load and unload cargo at specified freight centers (centrales de carga) and shippers must channel requests through these centers. While originally motivated by economies in provision of common services (such as provision of spare parts, insurance and paperwork), many freight centers began limiting entry and forming cartels in the assignment of cargo. Moreover, freight centers were increasingly reluctant to grant return-load facilities to other centers even on the same route, and opposed allowing vehicles from other routes to operate within their service area even when transport was in short supply. Centers at ports or border zones were even more restrictive because cargo could be easily centralised and controlled. 3 Monopoly provision of intenational container movements. Multimodal, a semipublic company granted monopoly status for the handling of international container traffic, was created by SCT in 1981 - with the exclusive right to provide door-to-door multimodal services. Collusion between Multimodal and the freight centers led to a system of surcharges and unofficial payments which often far exceeded the stated tariffs. Service standards were low and deliveries often delayed, resulting in a tendency for container traffic to be confined to port areas (and thereby foregoing the benefits of door-to-door delivery). In addition, customs regulations did not permit import-bearing containers that were temporarily brought into the country to transport domestic cargoes, resulting in empty outbound movement of containers. * Limited licenses for cargo handling. SCT granted concessions for handling cargo movements at railroad stations and border customs facilities as well as for the drayage services to cross the border. In most cases, goods could not be moved without the, services of the concessionaire. As with other concessions, entry was limited and cartels were formed. For private own-account carriers, permission was required to haul their own goods. 'Private fleets could not compete for back-haul and more broadly were not allowed to transport third- party cargo, again segmentation to prevent competition but in practice leading to sizeable unused capacity in the system. Eventually, a temporal permit allowed private truckers to offer 10 their services when public companies' supply was insufficient. However, private truckers were sanctioned with the loss of the truck if they offered the service wit;hout acquiring the permit. Companies that offer specialised freight services were regulated under separate rules. Specialised freight services included haulage of hazardous materials, objects of high volume or great weight, securities, industrial cranes and cars direct from the factory. These carriers required a different permission, but then could haul on all the federal roads. However, they could only operate the registered specialised trucks and transport a restricted nimber of specific cargoes (a subset of animals, perishable food, bottled liquids, construction and electrical material, furniture, automobiles). With respect to international competition, though cargo trailers were allowed to cross international borders and proceed to their final destination, the tractors (power units) were not permitted to cross borders, neither by Mexico nor by the United States. Drayage services have developed to undertake the exchange of power units at the border. The regulations for these services vary depending on the cities and states involved on both sides. In most cases the Mexican drayage firms are allowed to bring and take trailers across the U.S. border within a commercial zone l[imited to the urban area at the border. The U.S. drayage operations are generally more restricted as a result of Mexican regulations, at times not permitting border crossing nor the pick-up of return loads.20 An additional difficulty has been that the high- growth in-bond assembly (maquiladora) industries have: not been permitted to transport their own products in their own private fleets. Finally, foreign carriers wvere prohibited from owning Mexican trucking cornpanies. In additio:n to Federal regulations that affect interstate traffic, some states have restrictions that affect intrastate movement of cargo. In some cases, concessions are required for specific cargo and/or route movements on state highways. A review of state regulations in Chihuahua, Jalisco, Puebla, Veracruz and the state of Mexico has indicated that state regulations have not, in general, posed major issues as exclusive intrastate traffic is not very significant in comparison to interstate transport. Nevertheless, state regulations have been found to affect the movement of some agriculture, forestry and construction industry products.2' Market conduict. The cartel-like behavior stimulated by tight entry restriclions profoundly affected market ccinduict. Trucking companies especially small operators that offered similar services and loaded freight in similar locations were compelled to establish societies (lineas de concesionarios) and offer their services through freight centers. Some freight centers (like the ones in Mexico City and Guadalajara) reportedly restrained themselves to leasing spots to trucking firms that had direct contact with their clients, stamping the loading documentation (talones de emnbarque) and promoting with the authorities sanctions against those who offered the service without legal allowance. Other centers (like ihose in M:atamoros, Veracruz and Tampico) acted as inescapable intermediaries between truckers and shippers, preventing direct 20 For a detailed account of truck-related border crossing problerms, their persistence after regulatory reforms and their associated high transactions and congestion costs, see Giermanski (1995). 21 See World Bank (1990), p.14. 11 negotiation between carriers and shippers. They determined who carries a shipper's load, destroying carmers' incentives to build up reputation. Some freight centers compelled truckers to follow a roll call, first-come first-serve queuing system when loading freight. The fixed terms dictated by the freight center for the provision of the loading and delivery service were extremely generous: loading, one day; each 200 kilometres-segment transport, one day; unloading, one day. The sanctions for delays were restrained to partial losses of the rate, which favoured trucking companies: from I to 5 days, 5 percent; from 5 to 10 days, 15 percent, more than 10 days, 30 percent. This created incentives for excessively long delivery times, enabling the vehicle to wait a substantial time in order to find a return-load and not need to return empty. Thus, a trip from Mexico to the nearby city of Cuernavaca (a roughly 60- minute trip by car) could take 8 days with the shipper being required to pay 95 percent of the prescribed tariff. On the other hand, shippers who did not pick up their freight within the first 48 hours after cargo arrival to the destination had to pay warehousing charges.22 Important additional regulations or the lack of appropriate regulations affected vehicle maintenance and loading practices, vehicle fleet purchasing practices, as well as related safety and environmental practices. Main areas included: * Absence of dimension and weight regulations. Mexico lacked any regulation controlling vehicle weight and size. Axle overloading (excessive weight per axle) was suspected to be most serious on roads connecting major port areas, where cargoes are concentrated and heavy industries are located. Overloaded axles cause severe damage to highway pavements and bridges, which in turn increases the risk of accidents, limits road life and raises maintenance costs. - Absence of technical vehicle safety regulations. Mexico also lacked an inspection system to monitor vehicle safety. The present system allows potentially dangerous vehicles to use the public roads. At least 7 percent of accidents are due to vehicle factors such as bad tires and brakes. The proportion of number of persons killed on Mexican highways to the number of accidents was six times higher in Mexico than in Sweden.23 * Fleet purchase restrictions. The automotive industry, consisting of auto, truck, bus and parts producers faced numerous restrictions which greatly increased the effective costs of vehicle purchase and maintenance. For automotive production, the effective protection in 1989 was estimated at about 50 percent compared with a 15 percent average for manufacturing as a whole. In addition, imports of tractors and trailers were prohibited. * Absence of environmental regulations. Mexico lacked any norms controlling truck noise and air pollution. There were also only minimal programs in place affecting the quality of fuels. * Exemption from taxation. The Mexican transport sector was exempt from income taxes as well as value added and ownership taxes. Instead, since 1966, truckers were required 22 This account is based on Davila (1994), p.129. 23 The latter is based on a rough estimnate by World Bank staff. See World Bank (1990), p.17-8. 12 to pay a direct tax that was a fixed amount per operated vehicle. The armount varied according to the type of vehicle and service rendered, but was very small. The ecclusiorn provided a significant avenue for tax evasion since a manufacturer could start a trucking operation to escape corporate income taxes. Prices. Fixedjprces for routes. Official trucking tariffs were set by SCT, presumably above the corapetitive level and high enough to balance, on average, both fixecl and variable costs and yield "reasonable" profits. The rate was uniform for all seasons and for all regions of the country. l'he rates were classified in five different categories, each with its own variable factor per ton-kilometre. However all the categories shared the same fixed factor per ton. The classification complied loosely with the difficulty of handling different products, and it discriminated between the value of the goods. Trucking companies provided the cost inforr.nation that the authorities evaluated and used as a reference :For setting rates, but rates often bore little or no relationship with the costs of a specific shipment. Concessionaires were not allowed to chaLrge prices above or under the authorised rates.24 Neither could private truckers charge over or under the official rates when meeting extraordinary demand for the service.25 Additional application rates set values for other services such as pick-up delivery, exclusively served customers, empty back-hauls and warehouse spaces. Insurance liability limitations for carriers. Trucking companies' responsibi:Lities for damaged or missing freight were uniformly established for all the different kinds of cargo and were explicitly limited to very reduced quantities, unless the shipper paid an additional fee in proportion to the declared value of the good. The government had fixed the ceiling on the trucker's responsibiliity for cargo loss or damage to only US$0.3 l/ton. A further regulation prevented negotiation of incremental insurance by fixing the rate at which the shipper could insure his cargo irrespective of the product-specific shipping risks (at 3 mills per declared value). Non cost-reflective tariffs. While overall road costs were being recovered, there were important cross-subsidies between automobiles and trucks. By 1989, trucks were only paying for a small fractioin, roughly 15 percent, of the costs thali they caused to the highway network. This was essentially clue to the relatively low Mexican price of diesel fuel, which in late 1989 was roughly IJS$0.65;/gallon, insufficient to allow adequate cost recovery, coupled with the absence of significant license fees or direct taxes levied on truckers.26 2.2.2 Regulai'ory refiorm Market access. The basis of the deregulation program was that private investment and competition in trucking would be encouraged and the development of multimodal transportation would be pursued. The new access regulations included: 24 Some specific products such as corn, wheat, rice, sorghurm, sugar, salt, cement, coarse sand, gravel, sulfur, coal, raw oil and gasoline were subject to an 8% discount that functioned as a cross subsidy. 25 See Davila (1994), pp. 127-128 26 World Bank (1990), p. 19. 13 * Free entry into trucking with a simple permit. Any trucker wishing to operate one or more trucks need only prove his identity and vehicle ownership to get a federal trucking pernit. The changes abolished both limits on the number of concessions and on the partnership of incumbent trucking companies in the allocation process of new concessions. Specialised trucking firms are treated equally, allowed to transport any cargo whenever their trucks were technically suited for doing so. Private own-account operators are allowed to transport third-party cargo. All restrictions on backhaul are eliminated. * Free entry on all routes. Trucks have been freed to offer their services and transport interstate cargo throughout Mexico. Direct shipments between former corridors are permitted without additional costs. * Traffic can flow independently of freight terminals. Truckers do not have to belong to freight centers to obtain cargo and shippers are free to use the trucker of their choice. Freight centers no longer control cargo shipments nor certify the bill of lading. * Free entry into container handling. Entry into the provision of multimodal services has been opened up and containers can be transported without Multimodal's services. In January 1990, customs regulation restrictions on the use of international containers for the transport of domestic cargo were lifted. * Free entry into cargo handling. In January 1990, new regulations forbid giving exclusive rights for the provision of cargo handling services in federal areas, specifically at railroad stations, customs facilities and border crossings (drayage services). Any operator with a federal trucking permit is automatically allowed to provide such services. The current Regulation for Federal Cargo Trucking establishes the following requirements to obtain a permit: (1) fulfil the registration form that SCT issues (what it must include or maximum permissible items to include are not specified in the Regulation); (2) show evidence of third-party damages insurance; (3) declare the vehicle characteristics; and (4) show the certificate of low pollution emissions. For the case of incorporated legal entities (personas morales), the regulation also requires them to show the constitutive deed (escritura constitutiva) which establishes trucking service as their main activity (Arts. 7 and 9). The authorities' response to the application must not exceed thirty days (Art. 15). With respect to international competition, foreign carriers are still not permitted to make use of federal roads for international transport. NAFTA established a schedule of liberalization of the transportation sector in three, five and ten years. However, integration of systems between the United States and Mexico has been slowed by a U.S. unilateral decision to postpone NAFTA in trucking.27 Foreign carriers are not allowed to own Mexican trucking companies until 2004. One positive development is that January 1990 changes have permitted the maquiladoras to transport their own products in their own trucks across the border. 27 Based on arguments of divergent coordination in standards, lack of security in Mexican main highways and low technology of Mexican carriers. See Ibarra-Yunez (1999), p.6. 14 Althoughl the federal government developed a plan of action for federal-state harmonization of trucking regulation, the federal government has no jurisdiction over exclusive intrastate transport on state or local roads. Market conduct. Deregulation eliminated all restrictions on routes. It defined a new unitary route for public trucking that embodied all the roads of fedleral jurisdiction. Since it is no longer mandatory to contact clients through freight terminals, shippers have direct contact with their preferred trucking company and can make their choices on the basis of quality and price differernces. Otlher regulatory changes include: * Liberal dimension and weight regulations. Mexico introduced very liberal vehicle weight limits?W2Among the consequences are that Mexican roads need to be constructed at higher standards and costs than U.S. roads. In turn, Mexican vehicle operating; costs per unit of cargo are much lower as the higher road costs are not recovere(d in road user charges. On the other hand, regarding vehicle length limitations, Mexican standards are typically more restrictive than U.S. ones. In particular, Mexican standards do not allow the haulage of 51 feet trailers, the most common U.S. size. In practice, the size and weight limits are rarely enforced, due to the absence of sufficient platform-.scales. An insufficient number of inspectors also lack mobile umits and there is a reported lack of scales to enforce the weight regulations. This is an area where the authorities and trucking representatives of both eciuntries are still negotiating to reach a mutually acceptable agreement, under the guise of the post-implementation negotiations of NAFTA. If Mexican trucks are allowed to operate up to their own weight limit, Mexican operators would be at an advantage over U.S. operators and rail transport. On the other hand, if U.S. standards are adopted, Mexican operators will be disadvanl:aged as' their vehicles are designed to carry much higher loads and therefore have higher unit operating costs. * Liberalization of fleet purchases. The regulations governing the automotive industry were changed to ensure that truckers are able to purchase Units of international quality at international prices. Starting with model year 1991, the import of tractor-trailers was allowed if domestic prices exceed the international level, and for 1992 also heavy trucks. In addition, restrictive regulations on the organization of truck p:roduction, such as national integration requirements, compulsory incorporalion of domestic components into vehicles and the majority national ownership requirement have been eliminated. The privatization of the national truck producer DINA in the early 1990s and the market entry of aggressive truck producer 3 like Mercedes and V'olvo with financial lending leverage was critical. Level-playing field taxation. Within the 1990 budget approved by Congress, the exemption of the trucking industry from income, vialue added and ownership taxes was eliminated. Prices. The reform abolished official rates. The current regulation explicitly establishes that truckers rnust freely determine rates (Art. 65). These amendments fostered quality differentiation of the service, as rates reflecting differing quality were now freely negotiable 28 See World Bank (1994), Table 3.2, p.67. 15 directly between trucking service providers and users. As a complement, to protect shippers from cargo loss or damage, the regulation establishes that trucking companies will pay the equivalent to the declared value by the shipper (Art. 84). On December 1, 1989, the price of diesel was increased by 5.6 percent and this increased cost recovery to the highway network from trucks from roughly 15 to 50 percent. 3. Market Structure and Entry 3.1 Evolution of industry structure A dramatic increase in haulage units. Between 1989 and 1996, the number of haulage units increased from 164,010 to 315,318 (a 92 percent increase, or 13 percent per year, significantly higher than the economy average growth rate). The most striking increase was between 1989 and 1992, precisely concurrent with the regulatory reform period. As shown in figure 1, the total number of haulage units increased most dramatically between 1989 and 1990, by 36 percent, and by another 24 percent between 1990 and 1991. By the end of 1991, available capacity was more than 60 percent higher than in 1989. The total increase between 1989 and 1992 was of 85 percent (growth from 1989 to 1994 was 91%).29 What is also striking from figure 1 is that in spite of the 6.8 percentage fall in number of registered haulage units between 1993 and 1994, the number of units in 1996 had again surpassed 1993 levels, suggesting a relatively minor consolidation to date. This dramatic increase can be explained by: (1) the entry of new firms; (2) the expansionL of existing firms through acquisition of new haulage units; and (3) the legal registration of a number of formerly nonregistered firms (the transportistas piratas). As could be expected, the rigid regulatory restrictions not only stimulated collusion among formal truckers but simultaneously led to the development of a substantial informal trucking sector. This sector was comprised of small-owned operators (transportistas piratas) who avoided all legal constraints. Even though informal truckers could not take advantage of scale economies, they avoided indirect costs, did not pay taxes nor any other required fees. The transportistas piraw'as did not have offices or any other kind of administrative costs, and they secured their freight through informal freight forwarders (coyotes).30 As a result of easing entry barriers, many informal carriers decided on their own to register with SCT and avoid the risks and costs of operating illegally. In order to assess the extent of new entry as opposed to relabelling of existing trucks fro-m informal to formal status, it is important to examine the numbers in greater detail. By the end of August 1990, a total of 50,694 federal road permits for freight transport had been issued. Of this total, 30,153 (60%) were for new entrants, 13,830 (27%) for previously illegal operators. 29 In addition to being driven by the dynamics of regulatory reform, a portion of this increase may also be related to the prospect of NAFTA, even though in the early years its subsequent passage was far from taken for granted. 30 Even though one of the main outcomes of trucking deregulation was the inclusion of most of the informal truckers, some still subsist to avoid remaining regulations including taxation. 16 and the remainder (13%) for expansion of existing fleets.3' Based on this evidence, the most important group accounting for the increase was new entrants, suggesting that the amount of competition among truckers may have increased substantially. The combination of lower tariffs and improved service reportedly has induced some agricultural and industrial producers to give up their own in-house: truck fleets. They now outsource and contract their transport services from public providers. It is the elimination of some of these private fleets which has provided a significant portion of the physical vehicles for the expansion of the public for-hire fleet.32 A preponderance of small owner-operators. Under the preregulatory refarm period, in theory there were no 'large companies since no individual could be granted more than five vehicle franchises. In practice, however, some individuals controlled up to 500 vehicles through nominees. In any case, prereform official data vwas therefore not particularly meaningful. Bly 1996., the 315,318 haulage units were divided among 60,531 registered firms. Of this total number of firms, 90 percent were owner-operators (between 1-5 trucks), 8.3 percent were small firms (6-30 trucks), 2 percent were niedium firms (31- 100 trucks) and a further 3 percent were large firms (plus 100 trucks). Both the large number of owner-operators (54,292) as well as the relatively small number of large firms (187) are striking, as highlighted in figure 2. Of course, in terms of number of haulage units controlled by each size class, the large firm category becomes more significant, with almost 10 percent of units (9%). A dual segmented structure of technologically more advanced, concenirated large firms and low-technology, small fringe firms. Although large firms are relatively small in numbers, with 18 7 in 1996 (up from 148 in 1988), their market and contracting access and greater technological sophistication has reportedly maintained the system as a r elatively concentrated, dominant firm market structure. In 1994, the large firms with more than 100 trucks represented 65 percent of national haulage and 87 percent of transborder transport.33 The large firms provide services based on long-term contracl:s and cover main routes, while owner- operators and small firms cover short hauls, urban markets and cargo consolidation. Tentative evidence of th,e degree of segmentation is provided by evidence that: (1) large manufacturing clients of trucking finns generally do not use fringe carriers or subcontractors in case of nonanticipated demand; (2) large tucking companies generally do not consider lower prices of fringe companies as a signal of price competition; and (3) reputation, differentiation of service and quality are present in large trucking firms and not evident in the residual short haul cariers.34 31 World Bank (1995), p.29. 'Previously illegal operators' are those that indicated start dates prior to July 1989 but that were not previously registered; reporting was likely to be truthful as there were no ex post sanctions for such prior illegal activities though the number in this category may have beern substantially larger, at the expense of the 'new entrants' category. 32 Worild Bank (1994), p. 9. 33 As reported by Iba[rra-Yunez (1999), p.7. 34 Based on a survey of i13 leading trucking carriers (accounting for 26% of total road freight transportation) and 25 manufacturing producers users of trucking services in Northeastern Mexico in 1993 and updated in 1995. Out of the 25 industrial users, 18 did not use fringe carriers in case of non-anticipated demand, and 17 In response to the continuing foreign ownership restrictions, non-Mexican companies have to hire the services of domestic carriers who haul the trailers within Mexico. These regulations have fostered formal agreements and alliance between U.S. carriers and the larger Mexican trucking companies. Some of them prefer exclusive partnership arrangements (for example J.B. Hunt of U.S. has a partnership arrangement with TMM of Mexico) while others develop a. network of cooperative ventures (such as Contract Freighters Inc. that works with more than 20 Mexican companies).35 A relative scarcity of trailers. Another structural feature of the Mexican industry is the scarcity of trailers relative to motor units (both two and three-axle tractors and integrated trucks). While U.S. motor carriers typically operate 1.5 to 2 trailers for every motor unit, in Mexico, the ratio was 0.3 in 1989, rising to 0.5 trailers per motor unit by 1996 (figure 1). The obvious result for U.S.-Mexico trade is an overreliance on U.S. equipment in cross-border traffic. Put another way, the thousands of U.S. trailers operating in Mexico are there by necessity.16 This point is related to the fact that Mexican truckers are undercapitalised and they thus take every opportunity to operate without further capital outlays. In this context, it is interesting to note the increase in capital expenditures between 1989 and 1995 that underlies the rise in trailer to motor unit ratio over that period. 3.2 Industry conduct and intensity of competition Prereform low-quality equilibrium. As a result of the rigid prereform system, informal truckers found a market segment that was willing to take some risks in order to pay lower rates. It therefore appears that even under extensive regulation, the low-end of the Mexican market had a certain degree of competition, with illegal operators offering bargain prices. The biggest impact of the regulatory regime appears to have been in limiting the supply of higher- quality logistics services. For shippers that required better and faster service, it apparently could be obtained at a negotiated higher-than-official price. But even a larger trucking company could in general not position itself to consistently provide such a service due to peer monitoring and due to the impersonal system for matching loads and truckers. Consequently. no one invested significantly to improve quality, though there appears to have been some improvization to provide better quality service when paid to do so. The freight center system and CANACAR's intervention usually prevented shippers and carriers from developing long-term relationships where reputation and quality of service matters. As illustrated in figure 3, under regulation Mexico had a limited range of quality available, compared to the United States. Also note that rigorously enforced U.S. technical regulations would not have allowed Mexican-style low-quality trucking operations in any case. none of the 13 carriers used fringe company pricing as relevant to their own pricing decisions. See lbarra- Yunez (1999). 5 Supply Chain Directions for New North America, 1995. In addition to TMM-JB Hunt, other reported alliances include Celadon-Grupo Hermes, Carolina Freight Corp.-Tresguerras and Yellow Freight- Transportes Sierra; see Ibarra-Yunez (1999), p.7. 16ATA (1995). 18 Increased intensity of competition. Two types of measures -reflecting changes in intensity of competition were explored in the enterprise survey. A first question asked carriers to quantify for 1989 and 1997 the number of direct competitors whc, have or may have a significant impact on their business, asking them to report only the number of competitors who can take away significant chunks of business or from whom they can take away significant amounts of business. According to this measure, intensity of competition has increased significantly across the board, with carriers on average reporting an increase of roughly 50 percent in numbe:r of' such direct competitors. Intensity of competition has increased much more starkly for smaller firms (less than 30 trucks), who report an increase of over 85 percent, in contrast to large firms who report an increase of under 20 percent. A second question asked carriers to qualitatively rank five different classes of competitors (small owner-operators, large for-hire, private, affiliates of foreign carriers, othler transport modes) according, to intensity of competitive threat for :heir operations, both in 1989 and 1997. On average across al'l carriers, increases in competitive pressure were reported as coming from all classes of competitors. In terms of absolute levels for 1997, both small and large carriers report that the largest. competitive threat comes from large operators, followed by competition from affiliates of foreign carriers. Interestingly, in terms of changes between 1989 and 1997, the largest number oifcarriers reported increases in intensity of competitive threat coming from affiliates of foreign carriers. Both small and large firms report this finding. Small firms then report that the seconci most important increase comes from large operators, while large firms feel a stronger change in pressure from other transport mnodes. Only a minority number of carriers reported increases in intensity of competitive pressure from private carriers or from small owner-operators. After deregulation, prices for the existing quality of logistics services have fallen by at least 30 percent.37 Moreover, the NAFTA agreement has allowed large retailers and manufacturers to adopt faster and more complicated logistics systnems, forcing truckers to improve their services. This process is also helped along by U.S. truckers' interest to form alliances with Mexican truckers. Some U.S. finns work with different Mexican firms with only a formal agreement for trailer haulage. However, through a few more intensive alliances, some U.S. truckers are 1ransferring know-how, computer software, and equipment. As a result, higher quality trucking services are now available in Mexico. However, these services are offered by only a handful companies and therefore are cquite expensive (only albout 10 Mexican carriers own more than 500 vehicles). Consequently, the price-quality profile of Mexican trucking changed substantially (see figure 3). 4. Performance The followxing, analysis considers the impact of regulatory reform in road freight on industry-specific and broader economywide performance. Although it is always difficult to isolate the impact of policy from other changes in the economic environment, attempting to 37 Our surveys confirtned this widely believed figure. 19 quantify the importance of a specific set of policy changes is even more problematic during a time when so much else was changing simultaneously - including concomitant regulatory changes in related sectors, a drastic trade opening with Mexico's most significant trade partners, important related tax changes, and a severe economic downturn in 1995. In assessing downstream impact, therefore, we focus on the more modest objective of quantifying the maximum benefits for selected downstream users from key road freight logistics innovations. More generally, the detailed findings from the survey of both carriers and shippers supplement where possible and corroborate the trends in the official statistics. 4.1 Impact of reform on performance in the roadfreight industry 4.1.1 Output andprices Significant increases in output levels. Between 1989 and 1995 the traffic volumes in ton- km of domestic public road freight transport increased by 52 percent, from 107,243 to 162,827 million ton-kilometres. The average annual increase per year since reforms began, at 8.6 percent per year is more than double that during the 1980-89 prereform period (3.4% per year). Interestingly, the average distance carried increased by almost 30 percent, from 346 to 444 kni, after having been roughly constant for the previous nine years since 1980. A similar picture arises based on the volume of freight hauled. Between 1989 and 1995, there was an 18 percent increase in domestic road freight transported, from 309.8 to 366.7 million tons. The average annual increase since reforms began, at 3 percent, again is greater than during the prereform period (at 2.5% per year).38 In comparing these two series of output-related statistics, what is interesting to note is that although both increased significantly more during the postreform period, distances travelled increased substantially more than volumes carried, in effect longer trips to new and further locations. The survey data collected corroborates this significant increase in road freight haulage volumes since 1989. Of incumbent trucking firms reporting data for both 1989 and 1997, total annual kilometres hauled increased by roughly 60 percent. This increase masks important changes at the individual trucking company level, with some firms more than doubling distances covered, while two of the smaller firms decreased distances covered. A complementary indicator for freight haulage volume is changes in the number of trucks by firm. These figures show a clear redefinition of the market structure, with the growth of freight haulage volumes in the period mainly attributed to larger firms: * The average growth rate for trucking firms expanding their fleet was 125 percent: 9 o-ut of 14 (64%) reported an increase in their fleet; * The average growth rate for large firms was 175 percent: 7 out of 8 large firms reported an increase in the number of trucks, while only 1 reported a decrease; See IMT (1997), Tables 4.1.1-2. 20 * The average growth rate for small firms was 69 rpercent: 3 out of 6 small firns reported no changes in the number of trucks, only 2 of them reported an increase and 1 reported a, reduction. The number of states in which a firm has sizeable ciperations (deriving at l east 4 percent of its total income in terms of inter and intrastate traffic) rnay be a useful proxy for output mix, reflecting the transport of either existing or new products to new locations. According to this definition, trucking firms in the sample were active in between one and ten states, typically in two to five states. Based on this indicator, it appears that regulatory reform and liberalization of routes was associated with a broader range of services offered. The largest number of respondents reported an increase in the states that accolnt for an important portion of their operations, with 5 out of these 6 firms being in the larger size grcuping. Some of these expanding firms more than doubled the number of states accounting for sizeable operations, from four to eight, o;r from one to seven. The expansion in geographical area of operation is generally associated with the opening of one or more additional termninals or bases of operation. However, the fact that Mexico has 31 states and that the largest firm interviewed operates in only 10 of these suggests scope for continued consolidation of routes. Prices fall overall, but fastest in lower quality-of-service segments. Rate analysts in SCT found that between 1987 and 1994 trucking rates nationwide declined 23 percent in real terms. One official in SCT estimated that general cargo trucking cargo rates in 1994 on the major route between Laredo and Mexico City were about 30 percent lower in real terms than the prevailin,g rates i:n 1987.39 Another study concludes that while there is only incidental and anecdotal evidence on changes in truck tariffs, all the e vidence points to reduction of the order of 25 percent: in real terms.40 T he substantial reduction in overall tariff levels documented in available n1ationwide studies is corroborated by survey results. Almost all do wnstream users of trucking services interviewed reported that the cost of hiring a truck had fallen in real terms since 1989. Estimates of the size of the decline generally ranged between 5 anid 15 percent. More careful probing of aclditional exogenous factors confirm that there have been significant declines in real prices of a given. service delivered, though the magnitude of the price change is difficult to quantify given the variations in the actual service provided. For instance, one shipper estimated that an additional 20 percent price fall for the originall) available service should be attributed to the higher quality levels now available, including newer trucks, faster delivery, and more reliable shipping facilitated by more sophisticated tracking systerns. Another shipper estimated that the price fall would have been even more substantial if it did not incorporate the effect of new toll roads, which he estimated added 6 percent to the cost of a typical trip. Regarding rate structure, there appears to have been an important differentiation of prices to reflect diflferentiation of services provided to distinct classes of users. Customers who ship high value-to-weight components where timeliness and reliability of delivery are critical are 9 World Bank (1995), p.29. 40 World Bank (1994), p.8. 21 willing to pay substantially more for higher quality service. For such shippers, the logistics cost generally represents less than 10 percent of the product price. These services are generally provided by the larger, more sophisticated carriers. These prices do not appear to have come down as much, no doubt reflecting the increased quality elements embedded in the price and the relatively less intensive competition prevalent between the larger, technologically most sophisticated trucking fleets. On the other hand, customers who ship high volume, high weight products where the logistics cost is substantial generally seek to minimise transportation costs. These lower quality services tend to be provided relatively more by the small owner-operators. These are the prices that have come down most substantially. 4.1.2 Innovations andproductivity Large trucking firms adopt innovations. Based on survey results, a substantial amount of innovation in the industry has taken place with almost all significant 'better practice' logistics improvements introduced only after 1990. Importantly, the more sophisticated technological innovations have been almost exclusively adopted by the larger firms. All large trucking companies have modernised their fleet, with almost all buying new trucks with electronic combustion systems to minimise the use of fuel. Of firms adopting a standardised process for the purchase, maintenance and resale of trucks, over 80 percent were larger carriers. Most large companies also have taken advantage of computer systems to improve administrative controls and to upgrade their communications systems with customers through internet use. All large companies significantly modernised their freight tracking systems to ensure timely delivery, with a majority investing in modern satellite or cellular-based communications systems to institute some kind of tracking system. Note that a number of large truckers have U.S. based partners who urged them to introduce these innovations. Innovations introduced by large trucking companies since 1989 include: Technological truck-related innovations (e.g., new trucks with 7 out of 8 electronic combustion systems) (plus 1 in next 2 years) Irnproved computer systems and administrative controls 7 out of 8 Tnproved tracking systems 6 out of 8 (plus 2) Improved facilities 3 out of 8 Internet 3 out of 8 Human resources training 2 out of 8 (plus 2) ISO-900041 1 out of 8 (plus 2) Driver teams42 1 out of 8 Of firms realigning their routes along a reported 'modern hub and spoke system', again over 80 percent were larger carriers - such systems were typically introduced in 1993, 1994, 1996 and 1997. It is important to clarify that because most of the distances can be reached in less than 24 hours, the hub and spoke system as it works in the United States does not really 41 Truckers identified the ISO-9000 certification as an efficient way to gain contracts with U.S. trucking firms. Thus, this innovation could better be identified as an answer to trade opening. 42 Trucking companies believe that to maxiimize the life of the trucks it is better to assign the truck to one driver, who can be held responsible for it. They also mentioned that some drivers do not like to share theit trucks with other drivers. These are the reasons why driver teams are not considered very useful. 22 exist in Mexico. 'Based on the survey results, large firms generally focus on a few main routes instead of having a wide national coverage. The trucking firm with the highest coverage, for instance, provides road haulage services only in ten of the 31 states in the country. Large firms try to maxurLise their market share in the most profitable routes (e.g., Mexico City-Laredo, Mexico City-Guadalajara). What larger firms reportedly do is maintain loading and unloading terminals in the main cities, and switch from larger long-distance to smaller trucks to haul the trailers within the cilty. In this way, whenever a driver arrives to aL city he can switch trailers and continue working. There is no need to spend additional time in loading or unloading, and the trucking companies do not have to plan their trips i:n accordanice with clielnts' working hours. Generally., the trucks used for deliveries in the city have a higher average age. Accordinlg to a relevant industry study,43 the three leading LTL (less-than truck load) U.S. carriers now have full-time Mexican partners. On the urging of their U.S. partners, these LTL carriers adopted U.S. style management and operation procedures. This may partly be due to the fact that LTL was unheard of in Mexico before NAFTA and therefore U.S. firms had the opportunity to set up systems from scratch. Now LTL delivery standards are almost on par with those of the U.S., except that the Mexican LTL carriers do not have a large number of pick-up and delivery points which limits their volume, and they currently do not expect a big change in volume to enlarge their networks. Based on qualitative feedback during surveys, the incentives to introduce innovations for larger firms come from the dual imperatives under the new environment both to minimise costs in order to raise profilts and to improve service quality in order to gain new clients. As evidenced by the relatively low trailer to tractor ratios, Mexican truckers, especially the small ones, are undercapitalised. Following the economic crisis of 1994, most of them could not modernise or expand their fleets. This provided larger firms ample opportunities to invest to increase their market share. Larger trucking firms are reportedly also conscious that rail privatization will increase competition in the sector. Thus, the irrmprovements in quality are motivated by a desire to differentiate their service from both small trucking companies and rail carriers. Small truckiing firms struggle to survive. The contrast with small firms is extremely stark. Some reportedly have introduced no innovations whatsoever since 1989, while others have only been able to limit innovations to radio communications or limited improvements in office equipment.44 Only two out of the six small firms interviewed have been able to invest in fleet modernization and computer systems. Their planled innovations are more modest in scope, limited to modernization of existing equipment: * Only 3 out of 6 small firms introduced or were planning to introduce technology innovations in the next two years. 43 Supply Chain Directions for New North America, 1995 44 This finding of low ar no investment in technological upgrading by smaller firms is consistent with the findings of an earlier suirvey by Ibarra-Yunez (1999), where he concludes "reputation, differentiation of service or quality were not evidenced by the survey from the residual carriers of short hauls" (p.2 1). 23 * Small firms are more concerned to meet adequate spending levels for trucks' maintenance (3 out of 6). A few of them have trucks that cannot be used because they do not have the financial resources to afford the repairs. During the crises years they used old trucks to provide parts for the newer trucks. * A limited number have targeted improvement of the most basic administrative controls (3 out of 6). * One company was about to close because it faced stiffer competition that it could not meet, and it could not recover from the economic crisis. * Only one out of six interviewed was undertaking a serious effort to update the fleet, invest in computer systems and facilities. In general, small trucking companies do not have direct contact with their clients. They deal with freight forwarders (formal and informal) or larger trucking firms, and provide the service within a region. When they were asked how sector deregulation impacted on innovations, 2 out of 6 answered that higher competition in prices due to deregulation impeded the introduction of innovations. In the last years they have been mostly concerned about survival. Greater client responsiveness. Perhaps even more important than operational innovations, trucking companies introduced significant innovations in their relations with downstream users. This greater responsiveness to user needs has been the most widespread adopted new practice, instituted by over two-thirds of all surveyed carriers - and typically introduced in the years 1993, 1995 and 1997. Again, over two-thirds of the carriers introducing these behavior- type innovations were larger firms: * Exclusive transport contracts with large clients (adopted by 6 out of 8 large firms, but only 2 out of 6 small firms); * Cooperation with other carriers to improve return capacity utilization (4 out of 8 large firms, and 2 more will introduce it next year, while only 1 out of 6 small firms has such contracts); • Exchange agreements with other carriers to broaden geographical reach (4 out of 6 large firms, and 1 more will introduce it this year, while only 1 small firm has exchange agreements). * Specialised staff assigned to provide comprehensive logistics solutions rather than merely taking orders (6 out of 8 large companies, and 1 more will introduce it this year, while only 2 out of 6 small firms have specialised staff). * Standardised process for purchasing, maintaining, and reselling trucks (5 out of 8 large firms and 1 more is considering to introduce it, while only 1 out of 6 small firms has such process). Productivity on the rise. These innovations appear to have had a significant impact on productivity in the industry. Based on nationwide statistics, productivity as measured by output measures per employee increased substantially, as traffic volumes measured in ton- km increased by 8.6 percent per year since 1989 and measured in tons by 3 percent per 24 year, while employment increased at less than 1 percent per year (see below). Another measurable dimension of productivity change, the average age of the entire nationwide power units fleet (tractors for pulling trailers and integrated tractor-trailers, including an overwhelming majority of older owner-operated vehicles) has also improved, though only slightly with the average age falling from roughly 14 years in 1990 to 13 years in 1997. As shown in figure 4, purchases of new more technologically advanced equipment fell drastically during the debt crisis years of 1995-96, bujt have started increasing again. Projections of fuiture purchases and renovations anticipate the average fleet age falling to roughly 6 years over the next decade. 4 Survey resulits corroborate these nationwide statistbcs: * Total kilometres per truck have either stayed corLstant or increased between 1989 and 1997, on a carrier by carrier basis. Overall, total kilometres per truck increased by 42 percent and total kilometres per employee increased by 75 percent. * Average delivery times fell from 2.5 to 2 days (although the average distance carried was not controlled for in the question and may have varied somewhat between 1989 and 1997); * Average truck fleet age of large companies fell from 5.8 to 4.8 years. It rose from 6.1 to 9.7 years for small companies. 4.1.3 Profits, wages and employment Sharp falls in profit margins especially by small[er firms. Based on survey results, profits defined as pretax operating income as a percentage of revenues fell substantially in the industry. The fall in profits for smaller carriers was particularly severe, in the order of 50 percent between 1989 and 1997. Five out of 6 small firms reported that their profit margins had been reduced. Large firms also experienced falls in profit levels, but more modest ones in the order of 30 percent. Out of 8 large firms, only 3 reported reductions in profit margins, while 3 more reported no change and one actually reported an increase in profit margins.46 These results suggest that small firms on average had to sacrifice their profits to stay in the market, while aggressive large firms found in differentiating the service by quality an, answer to stiffer competition. Labor reallocation in favour of larger firms. Nationwide road freight statistics indicate an overall modest increase in employment in the industry betwe en 1989 and 1995 of 5.2 percent, from 509.5 to 536.1 thousand average annual remunerated employees (or roughly 0.9 45 Based on information elaborated by CANACAR from histc rical 1990-97 data from ANPACT and SCT. 46 These findings of low profits even for large firms may be at least in part driven by deterioration in macroeconomic ccinditions as a result of the 1995 debt crisis. Prior to that downtum, an earlier survey of 13 carriers found an average pre-tax profitability of 17% between 1993-94. 'In contrast, average profitability in the US sector of large carriers was close to 21% between 1989 and 1994. A similar substantial variation in profitability between large finns was found in that survey, with the leading carrier showing 12% growth in profitability between 1993-94 while the poorest performer registered a fa]l of 25%. See Ibarra-Yunez (1999), p.18. 25 percent per year). This increase masks a surprising increase of 9 percent over the initial two- year reform period, between 1989 and 1991, followed by relative stagnation between 1991 and 1993, an upturn in 1994 and then a sizeable decline in 1995 alongside the general macroeconomic downturn associated with the debt crisis.47 The survey results suggest that there was a substantial reallocation of employment in the industry. Incumbent small firms generally stay small, either contracting or expanding by a few workers. Among larger firms, less agile firms contract substantially, with one firm shedding three-quarters of its work force, from 1000 to 250 workers, while more aggressive firms expand substantially, expanding two, three or four-fold, with one firm growing from 150 to 720 workers and another from 180 to 447. 4.2 Impact of reform on performance in downstream industries The survey of users of trucking services focused on three broad types of industries as representative of different types of common users in Mexico: agroindustrial, electronics and auto-parts firms. In addition, two companies that specialise in the import and distribution of general goods also were interviewed. For purposes of analysis, it is useful to group the 15 interviewed companies into two classes of users: (1) high-intensity users of road freight services-shippers of higher volume or higher weight bulk products where the average trucking costs account for over 10 percent of sales, consisting of agroindustrial and import distribution companies (7 companies); (2) less-intensive users - shippers of lower volume or higher value-to-weight components where average trucking costs are less than 10 percent of sales, consisting of electronics and auto-parts firms, plus one producer of plastic parts for phones, printers and other electrical appliances. In the former category, trucking was always reported as the logistics service with the most important impact on costs, while in the second category air freight was frequently mentioned as more important though efficient trucking services were still deemed as essential for the firms' continued competitiveness. A set of more detailed follow-up interviews were conducted with two agroindustrial and two electronics companies in order to attempt to control more carefully for other exogenous changes in the operating environment facing these firms, as reported in separate boxes (based on the survey instrument of appendix 2). See appendix 3 for these case studies. 4.2.1 Costs Lower costs and higher quality. There are many factors that affect the inflation-adjusted cost of trucking services faced by downstream user firms. Especially among intensive users, substantial improvements in quality-including reductions in delivery times and transit losses-have positively affected the basic product provided, making an assessment of cost impact less straightforward: * Delivery times have decreased substantially according to 6 of 7 intensive users, with the impact directly linked to trucking deregulation. Better equipment and a reduction in the number of stops taken by drivers were cited as the key factors responsible for the 4 See IMT (1997), Table 1.1.4. 26 decrease. The average decrease was over 40 percent. Amorng less intensive users, delivery times decreased for 2 respondents, and remained unchanged for 4 out of 8. * Transit losses have also decreased substantially according to 5 of 7 intensive users, with higher q[uality equipment and higher levels cf security provided on federal roads (especially on toll roads) cited as the main factors. The average decrease was in the order of 35 percent. Among less intensive users, iransit-related losses also declined for 3 respondents, and remained unchanged for 4. * Reduced inventory levels have led to reported cost savings in 4 out of 1]2 responding firns, with all four being intensive users. More reliable and rapid transport services permit user firns to make more intensive use of both just-in-time delivery of raw materials, parts and components as well as build-to-order manufacturing for outbound shipmernts. In one agroindustry firm interviewed, in spite of increased sales, inbound inventory levels have fallen on average from 45 tD 18 days, while inventories of finished goo(ds have fallen from 7 to 3 days. * Direct truckiing costs were reported to have fallen. in real terns for 4 of the users, and all three are intensive users. Reported cost decreases related to trucking ranged from 33 percent to 15 and 5 percent declines. Upon more careful probing during, follow-up interviews, even those users that reported increases of 5 or 15 percent in real terms acknowledged the importance of reversing factors such as increases in the cost of toll roads or highier qluality trucking services as reflected in more rapid delivery times and significantly lower transit losses. 4.2.2 Innovations and new products New products - new goods, new areas and direcit deliveries. Since the onset of regulatory reformn, a number of intensive users have reported the ability to deliver new products to inarket as a direct result of cheaper or higher quality trucking services. 'New' products can include products that were previously not available on the market due to previously prohibitively costly or inflexible transportation services, as well as existing products thal can be economically shipped to new areas and new forms of delivering existing products (direct to customer rather than via wholesalers or retailers): * An expandecl range of distinct products has been offered by 8 of 13 respondents, includingo 4 of the 7 intensive users. As a specific example., a producer of fertilizer products reported increasing the range of producl:s sold from 12 to 32, with sales of new products accounting for 30 percent of total current sales. Of this increase 70 percent was directly attributable to lower cost and higher quality trucking services, with the remainder attributable to growth in demand due to the higher ]Level of fertilizer use in Mexican fields. As another concrete example, faster delivery times reportedly have helped a tequila producer expand the number of products (separalte brands) from 3 to 45, with average delivery times falling by 20 percent (half of this decline in delivery times is directly attributed to faster trucking services, with the other half attributed to faster loading and unloading at warehouses and ports). * Expansion of sales of existing products to new areas has been the main impact of improvements in trucking services for a number Df users, including 4 of 12 survey 27 respondents. A large Mexican producer of time-sensitive electronic components used by other downstream firns attributes 30 percent of expanded sales (which account for 50 percent of current total sales) to higher quality trucking services, with other explanatory factors behind the increase in new clients being Nafta-related tax changes and other internal logistics-related changes such as the introduction of bar code tags and computerised systems for faster and more reliable documentation handling. As another example, sales in new areas as a fraction of total sales of a company specialising in the production and distribution of Mexican peppers account for 90 percent of total current sales. Improvements in the quality of trucking services, with losses from transit-related problems falling from 5 percent of sales to 4 percent of now- higher total sales and a 10 percent reduction in delivery times, have reportedly played a significant role in generating these expanded sales. Direct delivery to retailers or end-users has allowed some user companies to earn higher net margins. Based on a more in-depth set interview with an agroindustry firm, the gain from direct delivery was in the order of 0.4 percent improvement in the firm's operating margin, with the benefit arising from the joint impact of new end-user customers that receive direct deliveries and the higher net margin on such direct sales (after controlling for other factors that could also have accounted for increased direct deliveries). More efficient logistics systems. Cheaper or more customer-responsive trucking services can and reportedly have led to a number of other logistics-related innovations within user finns: * Centralizing manufacturing, using cheaper or more efficient trucking to concentrate manufacturing in fewer locations has led to reported cost savings in at least one firm. Centralizing warehousing, concentrating warehousing operations in fewer locations also has had a similar cost saving impact. The joint impact of both factors, controlling for other factors such as changes in demand and rail shipping, was in the order of a 0.6 percent improvement in a specific intensive user's operating margin. * Rationalising loads, both through adopting standard lot sizes and containerization, has led to further cost savings for some user firms. Containerization, in turn, both contributes to cheaper transport and can play a role in decreasing damages and losses during transit. A number of interviewed firms have reported costs savings from these logistics-related changes, which in turn have generally been introduced only in the posil deregulation period. * Yet another example of improved logistics services has been the emergence of two large integrated logistics companies, each with about one million square feet of warehousing space. 4.2.3 Additional economywide impacts Outsourcing and enhanced rivalry. For a variety of infrastructure services essential for entrepreneurial activity, the absence of a well-functioning market in such services requires firms to establish in-house capabilities or forego production. For road freight services, the 28 alternative to purchasing transport services responsive to specific user needs on the open market is to invest and maintain a private own-account fleet of trucks. From an industrywide perspective, such required fixed cost expenditures may be prohibitive for cash-starved new entrepreneurs. Altholugh such costs are not largely sunk, tapping capital markets may not be an option for prospective entrants lacking reputation. An imnportant additional benefit of a more competitive trucking market, therefore, is to allow firms to outsource their transport requirements, in e ffect converting a fixed cost into a variable cost. The benefits from enhanced new entry and rivalry for markets intensive in the use of trucking services are likely to be substantial. Other benefits. As a result of reductions in truck tariffs, SECOFI has estimated that the overall distri'bution costs of commodities in Mexico has declined about 25 percent in real terms between 1987 and 1994. Deregulation of trucking also had an impact on the then-publicly owned railway company, Ferrocarriles Nacionales de Mexico (FN4M). The reduction of trucking rates after 1989 enabled truckers to compete miore effectively with FNM, causing diversion from the railway to truck transport, particularly during the period 19I89-91 when rail traffic declined si,gnificantly. 5. Conclusions and Policy Options for Reform 5.1 Successes andf,4ilures of reform The Mexican trucking sector is going through a radical transif'ormation following deregulation and Mexico's entry into NAFTA. Regulatory reform should be credited with increasing competition. Its impact was smallest in the lower-end of the market that was already competitive with a large number illegal owner-operators. Mid- and large-size shippers, however, hadL a chance to take advantage of deregulation to reduce their trucking costs. With the overall fiall in trucking prices, inefficient operators with inadequate equipment have been exiting the market. Regulatory reform also had an impact at the upper-end of the market where shippers reqirire faster and more reliable service and are ready to pay for it. Under the former rigid regulation regime, no carrier had the incentive to offer high quality services. However, NAFTA gave a bigger boost to this market segment. VWith closet trading ties to the United States, Mexican firms had to upgrade their logistics operations. Large carriers, often as partners of U. S. trucking companies, have begun to of ter improved services. Main lessons of Mexico's experience for other coLuntries include: The positive role of increased competition in the road freight industry in fostering economywide innovation and growth. Besides expected gains from lower trucking prices, the removal of access barriers to transport infrastructure matter's by generating substanLial downstream benefits, facilitating entry and new products. D)elivery times and transit losses have decreased substantially. More reliable and rapid transport services hav e permitted user companies to reduce inventory levels and make more intensive use of both just-in-time delivery of rawv materials, parts and components as well as build-to-order manufacturing for outbound shipments. More irnportantly, faster and more reliable trucking has allowed user comipanies to offer new goods, both 29 introducing previously unavailable products as well as making it possible for existing products to reach new areas. More customer-responsive trucking services have reportedly also led to a number of other logistics-related innovations within user firms, such as centralizing manufacturing and rationalising loads. Finally, some user firms have outsourced their transport requirements, in effect converting a fixed into a variable cost and thereby significantly lowering entry barriers. Successful reform requires careful planning, execution and high-level political support. A project with the objective of reforming existing regulations and introducing increased competition must devise a strategy to carefully take into account key interested stakeholders, both supporters and opponents. In the late 1980s, transport regulation in Mexico was viewed with growing disdain, particularly as a result of studies documenting the high economic costs associated with the prevailing system. The concurrent general program of structural reform was a favourable context for regulatory reform of the trucking industry. Arguably even more could have been done in the early stages of implementation to disseminate the costs of the existing system and the expected economywide benefits from increased competition. Given the potential for blockage of reforms by the trucking chamber and its members that had been benefiting from the officially sanctioned cartel, certain key steps in Mexican planning and implementation include: preparation of the initial reformn by a small group independent from SCT (where certain officials who had traditionally had clientelist relations with truckers or who feared losing their positions could have sabotaged the initiative early on); presentation of the reforms to recalcitrant truckers as a 'modernization' effort rather than an antioligopoly competition thrust, with the promise of soft loans for fleet modernization in exchange for the charnber's non-opposition; a sequenced deregulation, eliminating tariff ceilings with a lag following the elimination, of entry restrictions to prevent a possible consumer backlash if the forces of competition would have been initially too weak to prevent sharp price rises; unconditional support from the highest political levels, in this case from President Salinas himself. - Sufficient attention and resources should be devoted to assist the oversight institution in adapting to postregulatory reform conditions. Regulatory reform causes profound changes in the sectoral institution of government that formerly had responsibility for the regulation. Any regulatory reform effort should assign sufficient resources to assist the organization in making the transition to postregulatory reform conditions, assisting with such tasks as defining the organization's new role, reorganizing its structure and facilitating redeployment of staff. * Promarket rules to offset remaining market failures should be introduced concurrently with removal of distortionary anticompetition rules. Ideally, any new regulations to be introduced as appropriate complements to a more market-oriented framework (such as programs of highway safety, control of vehicle weights and dimensions, and inspections of vehicle emissions) should be implemented at the same time as the deregulation actions. In this way, there can be a smoother transition from the government-controlled to the postreform market-determined regime. * The competition agency has a critical role to play in any regulatory reform initiative, both in terms of up-front advocacy and ex post enforcement activities. The national 30 competition agency should play a critical role in hlelping lay the groundwork for reform, bym aking as compelling a case as possible for the costs of the prereform regime and for the expected benefits of refonn, through media relations and more systematic relationships with representatives of consumer and producer interests (chambers of commerce, small business and entrepreneurship organizations, consumer protection institutions, consumer groups, local training and research institutes and other relevant interested parties). In addition to this prereformn advocacy and education role, the competition agency must play a careful postreform enforcement role to ensure that cosy cartel-like behavior stimulated by tight entry restrictions does not persist. It is very natural lthalt public policies of segmentation and control of prices and quantities leave a strong iraprint on postreform market conduct. Inefficient, anticompetitive public regulation should not be allowed to be repLaced by socially inappropriate anticompetitive interfirm agreements. The experience of Alexico is instructive in this regard. At least three separate strong interventions were required by the Mexican Federal Competition Commission to discipline attempted anticompetitive practices by the trucking 'industry in the years following the irLitial regulatory refomis. 5.2 Potential beniefits and costs offurther regulatory reforms An interesting indicator of the need for further reforms can be derived from the perceptions of providers of road freight services themselves. Two questions wiere explored in survey format, a first on renmaining problems facing the indust-y and a second on features of the current legal and regulatory framework that restrict operations or inhibit growth. When questioned about remaining most important problems facing the road freight industry, trucking companies were near unanimous in highlighting the security problem, including robberies on highways and additional required payments and costs, as the number one outstanding policy issue (1 1 out of the 14 trucking companies surveyed idlentified it as one of three of four key issues, with most listing it first). The second issue in order of importance was the continuiniig pressure felt by increased intensity of competition. This issue was reflected in answers ranging from excessive competition from abroad or from the informal economy by larger carriers, unfair competition by smaller carriers (presumably from owner- operators in the informal sector who avoid all taxes), and as a problem of access of credit by some small and larger carriers (who presumably cannot raise finance at suffic:iently attractive terms given rates of return in the sector). At least one large carrier reported that they did not want more foreign firms to enter given their larger size, scale economies and better financing, which suggests still insufficient (at least international) competition. A number of the carriers emphasised ihat 1hey do not have access to loans at the same low rates of interest as their U.S. competitors. Interestingly, the multimodal freight broker surveyed highlightecl problems of timeliness of delivery and delays in service as the most important problem, suggesting that competitive pressure still had not been sufficient to stirnulate broad service satisfaction.48 Finally, the third issue raised most frequently concerns vehicle roDad safety, vwith the issue 48 This view is supported by the Ibarra-Yunez (1999) survey, where 17 out of 25 industrial users responded in the negative to the question of whether users find supply of iroad freight service adequate in volume, price- competitive, complete and dependable (p.20). 31 arising either under the heading of excessive number of accidents, better required control of drivers, or poor condition of highways (6 of the respondents highlighted one or more of these topics, with it being one of the two issues raised by the multimodal freight broker as well). When asked about the more specific question of key features of the current legal and regulatory framework that restrict operations or inhibit growth, the most frequently reported issue related to the weights and dimensions rules (5 out of 11 respondents). Answers ranged from absence of rules, lack of clarity of rules to inconsistent application (allowing old, too heavy trucks to remain on the roads in the case of a modem fleet owner, preventing old, too heavy trucks to circulate in the case of a struggling, nonmodernised company). The second most frequent issue was the relatively high level of per-truck license payment (tenencia) now required in order to operate a vehicle, allowing closer to full-cost recovery of highway maintenance expenditures. Other issues mentioned by more than one carrier include remaining ambiguities in the foreign investment law, and shortcomings of the labor law. Interestingly, the only legal/regulatory issue raised by the multimodal freight broker was problems with remaining labor unions in ports, suggesting that prior legal and regulatory barriers in trucking have largely been addressed. Further improvements in Mexican logistics will materialise only through closer links to the U.S. truckers. Currently the truly international market segment is populated by a few Mexican trucking companies that have established different forms of partnership with U.S. carriers. This segment is probably awaiting year 2004 when U.S. trucking companies will be allowed to own Mexican companies. If and when Mexican truckers are allowed to operate into the United States, it may inject a boost to productivity improvements. These developments will benefit mainly those parts of the Mexican economy that operate internationally. As far as logistics is concerned, there is an enormous amount of know-how and experience that Mexican companies could benefit from. Moreover, Mexican companies could benefit from improved logistics services without incurring substantial restructuring costs (unlike, say, taking advantage of equity markets in the United States). There are three main factors that hold back a full-scale logistics revolution in Mexico: * The persistent segmentation between large, technologically sophisticated providers o f road freight services and low-technology fringe providers. The parts of the broader Mexican economy that are not operating internationally still remain much less developed, relying on informal establishments for distribution and retail sales. This part of the economy cannot afford and does not seem to need improved services. Luckily for them there is a large fringe of small owner-operators to provide haphazard but inexpensive transportation services. In fact, the existence of low quality but cheap trucking is an important lifeline for small, informal businessmen. However, this fringe exists at a bigger cost to the rest of the economy. Their technically inadequate trucks are a threat to road safety, and their inadequate consolidation continues to shelter the larger carriers from more intense competition. * The Mexican govenrnent's unwillingness to enforce technical standards and maximum load restrictions. By refusing to implement technical standards rigorously, the Mexican 32 government is allowing the informal sector to can.Ty on. It is allowing a dual structure to persist that hinders more systemic modernization and more homogenous growth. Better enforcemnent of technical standards in trucking as well as in other areas of the small business enviroiLment will no doubt put some informal businessmen oult of work, but the ensuing consolidation could very well create new opportunities of greater benefit to the economy. * Failure to full implement the transportation related provis.lons of NAFTA and to ensure more ]nomnogeneous federal-state regulatory reforms. Although pressure from increasecd intensity of competition is being felt in Mexico, there appears to be scope for further significant gains from additional competition that a full opening with the United States could provide, as suggested by users' emphasis on continued service quality shortcomings. U.S.-Mexico integration has lagged behind the needs of the market, with negative consequences for Mexico's international competitiveness. In addition, failure to ensure more homogeneous federal-state regulatory refon:ns risks undermining progress to-date in the areas of competition, environmental standards anid road safety guidelines. Progress on these remaining internal and international obstacles could yield substantial additional benefits. Indeed, answers to a final question posed to users of road fLreight services suggest scope for significant further economywide benefits from increased trucking service efficiency. Under the hypothetical scenario that transport and storage costs could fall by a further 50 percent, the market value of user firms was estimated by respective managers to an average of 20 percent with some firms estimating up to a 100 percent increase. Under the hypothetical scenario that the frequency of deliveries to clients could double without additional charges while maintaining the volume of sales constant, similar substantial gains to the value of user enterprises were reported. 33 References ATA (American Trucking Association), 1995, South of the Border, 2nd edition. A.T. Kearney, 1998, Survey of Supply Chain Management, 1998, as reported in Financial Times, Supply Chain Logistics, FT Survey, Dec. 1. Comisi6n Federal de Competencia, 1996, Infonne Annual de Competencia Econ6mica 1995-1996, Mexico City, Mexico. Davila Capalleja, Enrique Rafael, 1994, "La Reglamentaci6n del Autotransporte de Carga en Mexico," in Francisco Gil Diaz and Arturo M. Femrndez (eds.), El Efecto de la Regulaci6n en Algunos Sectores de la Economia Mexicana, ITAM-CINDE-FCE (Lecturas de El Trimestre Economico No. 70), Mexico City, Mexico. Davila Capalleja, E., 1995, "El tratado Trilateral de Libre Comercio en materia de transporte de carga'', Chapter 9 in Lo Negociado del TLC Femrandez, Arturo, 1993, "Trucking Deregulation in Mexico" in Jose Carbajo (ed.) Regulatory Refomri in Transport. Some Recent Experiences, A World Bank Symposium, Washington, D.C. Giermanski, James R., 1995, "Texas to Mexico: A Border to Avoid", mimeo, Texas A&M International University (September 26, 1995). Ibarra-Yunez, Alejandro, 1999, "Dominant firm and company integration: The case of transportation in Mexico and the U.S.", Latin American Business Review, Vol.2, No.4, USD-ITESM, International Business Press. IMT (Instituto Mexicano del Transporte), Secretearia de Communiciones y Transporte, 1997, Manual Estadistico del Sector Transporte 1995, IMT, Sanfandila, Queretaro. Islas, Victor 1990, Estructura y Desarrollo del Sector Transporte en M6xico, El Colegio de Mexico, Mexico City, Mexico. Martinez, Gabriel and G.Farber, 1994, Desregulacion Economica (1989-1993), Fondo de Cultura Economica. Moore, T. Gale, 1975, "Deregulating surface freight transportation" in A.Phillips (ed.), Promoting competition in regulated markets, The Brookings Institution, Washington, D.C. Murray-Lasso, Marco Antonio, 1994, "Canales de Comercio: una Estrechez Comprometedora," in Gabriel Martinez y Guillermo Farber, Desregulaci6n Econ6mica (1989-1993), Fondo de Cultura Econ6mica, Mexico City, Mexico. Porter, Michael, 1997, "What is Strategy?" Harvard Business Review. Rico, Oscar, 1996, "Algunos Indicadores de la Evoluci6n de las Empresas de Autotransporte de Carga en Mexico, 1988-1993", in Instituto Mexicano del Transporte, Notas, num. 31, Quer6taro, Mexico. Sable, Charles, and Michael Piori, 1990, The Second Industrial Divide, Basic Books. Secretaria de Comunicaciones y Transporte (SCT), 1996, Estadisticas Basicas del Autotransporte Federal, Mexico City, Mexico. Sempe Minvielle, Carlos, 1997, Tecnica Legislativa y Desregulacion, Editorial Porrua, Mexico City, Mexico. U.S. Department of Transportation, 1996, Census Surface Trade Flow Data. Washington, D.C. Viscusi, W.Kip, John M.Vernon and Joseph E.Harrington, 1995, Economics of Regulation and Antitrust, The MIT Press, Cambridge, Mass, 2nd edition. World Bank, 1995, Perforrmance Audit Report - Mexico: Road Transport and Telecomnmunications Sector Adjustment Loan, Report No. 14400, The World Bank, Washington, D.C. (April 21, 1995).. World Bank, 1994, Mexico: The end of transition - A review of the transport sector, Report No. 12654- ME, The World Bank, Washington, D.C. (May 31, 1994). World Bank, 1993, Project Completion Report -- Mexico: Road Transport and Telecommunications Sector Adjustment Loan, Report No. 12623 (December 23, 1993). World Bank, 1990, Report and Recommendation of the President - Mexico: Road Transport and Telecommunications Sector Adjustment Project, Report No. P-5254-ME (May 3, 1990). 34 Appendix 1. The Road Freight Industry In Mexico Mexican GDP per capita is about 10 percent of its main trading partners, the United States and Canada. With a skewed income distribution, more limited car ownership, high degree of concentration in urban areas, and persistent poverty in rural areas, the Mexican economy has somewhat difftrent transport and logistics needs than its North American neighbours. Mexican industry definitions. Road freight services refers to the movement by trucks of goods (raw materials, capital goods, processed inputs, finalL outputs) for business purposes on the national road system (ranging from rural feeder roads, urban roads, to state and nalional highways), transforming a specific good at a certain location and time to the same good at another location and time. Road freight services are the most important of all combined freight and passenger transport services provi(led in M:exico, accounting for 40 percent of gross value added of the transport and storage sectors in 1995. 49 Road freight haulage units include power units (tractors used to pull trailers as well as integTated tractor-trailers, generally either two or three-axle) and separate trailers. Services offered (depending on the haulage units and goods moved) are classified either as general cargo or specialised cargo movements. Specialised cargo currently includes seven categories: hazardous materials, petroleum and its derivatives (via liquid tanker trucks), automobiles direct from the factoiy (via specialised vehicles), tow-trucks, tow-trucks for salvage vehicles, industrial cranes, and securities of credit-related institutions (via armnoured vehicles).0 In Mexico, road freight services are further classified according to three criteria. The first criterion derives directly from the Mexican regulatory distinction betweenfederal and state trucking companies, according to whether truckers have been granted permission to operate at the federal level or whether operations are restricted to a single or several states. Most of the available data omits information on firms and individual haulage units that are registered exclusively with state authorities.51 HIoNwever, this is reportedly not a major problem since exclusive intrastate traffic is minor in comparison with interstate transport. The seconid important criterion is between public (third-party, for-hire) trucking companies andprivate (in-house, own-account) trucking fleets. Again, most of the available data omits information on the size and iimportance of fleet trucks. About two-thirds of total cargo is transported through public trucking 2 A third useful distinction regarding the structure of the industry is between owner-operators (entrepreneurs operating between one to five trucks) and larger trucking firms. Total recorded federal public trucking in 1996 consisted of 315,318 haulage units (two-thirds of which were power units). These haulage units were operated by 60,531 registered firms, of which 89.7 percent were owner-operators. General and specialised cargo services in terms of volume transported represented 89.6 percent and 10.4 percent of the national total, respectively. The states wvith the largest number of haulage units 4 See IMT (1997), Tables 1.1.1-2. Total transport and storage services, in turn, accounted for 7.6% of gross value added in Mexico in 1995. 50 Perishable goods (via refrigerated trucks), goods of high volumne or great weight (via specialized platform vehicles), electrical and telephone equipment, construction materials, glass, and agricultural non-processed goods and animals are all examples of products that were previo usly classified as requiring specialized services but are considered as part of general cargo since 1991. 5 SCT, the federal Ministry of Communications and Transport is currently umdertaking a rnassive effort to re-register and re-tag the entire national trucking fleet. That effcirt will result in a comprehensive registry including holders of state-only licenses, but is expected to be comnpleted only later in 1999. 52 World Bank (1993), p. 7. 35 are Distrito Federal (22% of the total), Nuevo Le6n (10.5%), Estado de Mexico (7.4%) and Tamaulipas (7.3/%). 53 A potentially highly competitive industry. The markets for road freight services are quite fragmented because of the multitude of destinations (local, domestics, and international), types of goods to be transported, and multi-modality requirements. Unfortunately, the potential competition between road and rail freight services has been minimised in Mexico by the official tariff setting policies, the rigid radial structure of tracks and the un-addressed operational problems of the state-run rail company. As a result of official tariffs, railroads mostly carry bulkier, low-value goods, such as raw materials. In 1995, grains accounted for 20 percent of tons transported, cement 15 percent, primary steel inputs for 10 percent, non- metallic minerals 9 percent and diesel fuel 7 percent (ointly over 60%). In contrast, road freight services overwhelmingly transported manufactured or processed goods, with processed industrial goods accounting for 49 percent of daily tons carried, agricultural goods 16 percent and various packaged goods a further 14 percent; with unprocessed mineral goods accounting for 4 percent.54 Moreover, the initial rail network created a rigid radial structure, with all major routes passing through Mexico City. Finally, inefficient routing, maintenance and repair operations of FERRONALES (Ferrocarriles Nacionales de Mexico, the national railway) led to substantial delays, under-utilised capacity, and frequent non-availability of scheduled train services.5 The dominance of road freight in transport services. It is therefore not surprising that road freight transport increased its dominance over rail after deregulation. In 1989, road freight transport accounted for 65 percent of the total and rail 24 percent; the corresponding land surface transport modal split between road and rail freight was 74:2656 By 1995, road freight increased its share of domestic ton-km to 74 percent. Average road freight distance was 444 km (of a total of 220.4 billion ton-km of cargo transported). In contrast, rail accounts for 17 percent (average distance of 717 km), maritime transport for 9 percent (630 kin), and air transport for a mere 0.03 percent (862 km).'7 The total movement of cargo over the years has been influenced by the general growth of the economy, with total cargo transport growing at an annual rate of 1.7 percent per year over the 1980-95 period in spite of a period of stagnation during the mid-eighties. Across modes, road freight increased at an annual rate of 2.5 percent, rail fell by 1.8 percent per year, while maritime and air transport had only very minor changes (maritime growing at 0.5 percent per year, air actually declining at 0.4% per year) 58 In terms of the important U.S.-Mexican trade flows, trucks carry 77 percent of trade in value terms, rail 9 percent, maritime 10 percent and air transport 4 percent.59 5 These statistics are from SCT (1996). 54 See IMT (1997), Tables 4.3.5 and 4.2.5. It is noteworthy that the concentration of rail transport on a few raw materials has not changed substantially over the past years. In 1987, grains accounted for 24% of ton- km, primary steel inputs 19%, non-metallic minerals 10% and cement 9%, as reported in Davila Capalleja, 199x, p.300. 55 Ibid, p.296. 56 In contrast, the road-rail freight modal split in the US in 1989 was 41:59. 57 Road freight only captures public transport and excludes private in-house while rail excludes cargo moved for own purposes. Since the former surely is more important than the latter, these numbers underestimate the importance of trucking in freight transport. s See IMT (1997), Table 4.1.1. 59 US Department of Transportation (1996). 36 Infrastructure shortcomings. Despite heavy reliance on trucking, Mexico has a sparse and inadequate highway system of only 48,087 km of paved federal roads, referred to as trunk highways (com.pare to 'Canada's 800,000). Of this total, roughly 6,000 km are so-called concession highways, which are privately owned toll roads; this is a recent development, as concession highways only accounted for about 1,000 km in 1989.60 Although a welcome development, these new highways are grossly under-utilised, mostly because they are very expensive (a trip from Nuevo Laredo at the U.S. border to Acapulco can cost a trucker about US$400). The national system is deepened by an additional 46,000 km of paved state roads, r eferred to as feeder roads.6 Broader transport reform initiatives. An attempt is currently under way to address the traditional under-utilization of rail in Mexico. As part of a recent effort to restructure and modeniise the Mexican rail system, the Constitution was reformed in February 1995 to open the sector to private participation through the awarding of long-term concessions on groups of lines. A subsequent May 1995 law has created the legal framework for privatization. Licenses to build and operate railways were granted, with the national grid having been divided into three main vertically integrated regional companies. The Northeast line (Ferrocarril del Noreste), one of the country's busiest, was sold to a Mexican-United States consortium ir, January 1997. The Northwest section (Ferrocarril del Pacifico-Norte) was sold in August 1997, and the Southeast line (Ferrocarril del Sureste) in June 1998. Interconnection services in the Mex:ico City valley (Terminal Ferroviaria del Valle de Mexico) are provided by a separate company joinl:y owned by three big railroad lines. This structure encourages competition within the rail network via regional comparisons of costs and prices. Competition with road transport is promoted by eliminating the possibility of cross subsidies between regions and by stimulating the efficiency of the network. Mexican ports also have been plagued with problems cif inadequate equipment, lack of warehouse or terminal space and poor road and rail connections. Mexican ports have been decentralised since 1992, with port management -includin o negotiation of wharfage rates directly with steamship lines under maximum rate guidelines-- fonnally transferred to newly created Integrated Port Authorities (APIs) in the second half of 1994. Concessions for container and other handling services at three of the country's four major ports (Altamira, Lazaro Cardenas and Manzanillo) were sold during summer 1995. Other port concessions, including handling services in Ensenada and further such services in Lazaro Cardenas and Manzanillo were sold in 1996-8. Finally, the aiiport infi-astructure sector also has been operled up. The sale of concessions to manage the 35 (of 61) airports that do not require subsidised operations has been organized into four regional groupings, with the sales expected to be completed during 1999. 60 The govermnent handed over the construction of all four-lane toll roads to the private sector in the early 1990s. 61 See IMT (1997), Table 6.2.2. 37 Appendix 2. Structured Survey Instrument: Fertilizer Company Example Measurement Actual Fraction of Definition of Were there figure net change due the impact signiricant of to the What to variable Actual reversing reversing specified Impact factor I Impact on Innovation' Impace2 Measure? (historical data) figure factors? factors innovation (current data) net margin TRUCKING Cheaper or Reduce cost of Cost of hiring a Percent 0.15 If the trucker 0.21 1 Trucking 0.21 0.0441 higher transportation truck on a reduction in cost do not use toll costs/sales quality standard route of hiring a truck roads, they trucking since 89 (in real may offer a terms or U.S. further dollars) (hypothetical) discount of 6 percent. Introduce new Sales of new Sales 97 of new 0.3 None 0.3 0.7 Net margin on 0.07 0.0147 products products (incl. products new prods. - discrete quality introduced since average net changes) 89/ total sales margin Cover a wider Sales in new Sales in new 0.3 None 0.3 0.7 Net margin in 0.04 0.0084 geographic area areas that are areas (reached new areas - (begin selling in new reached by trucks 89-97) I sales average net regions or countries 97 margin reached by trucks) Make direct Number of new Number of new 2 NO 2 0.6 Net margin for 0.01 0.0036 deliveries to more customers (who direct-delivery new customers customers start buying from customers / total - average net you after 1989) number of margin who receive direct-delivery direct deliveries customers in 1997 Direct delivery to Percentage sales Increase in -0.2 Their 0 0 Net margin in 0.02 0 retailers or end volume switched (sales to marketing sales to retailers users allows from wholesellers retailers) / (sales strategy focus - net margin in company to eam to retailers (or to wholesellers) in increasing sales to higher net margins large customers) from 89 to 97 the proportion wholesellers of whole- sellers. Increase in 0.03 Their 0 0 Net margin in 0.02 0 (sales to end- marketing sales to end- users) / (sales to strategy focus users - net retailers) from on increasing margin in sales 89 to 97 the proportion to retailers of whole- sellers. Higher quality Total transit loses Reduction in 0.01 None 0.01 0.8 1 0.008 trucking services = uninsured total transit limit losses (theft & losses + loses/sales from damage) in transit insurance 89 to 97 premiums + loss prevention expenses Centralize Concentration of Percent change 8 None 8 0.5 Fixed 0.01 3 0.0052 manufacturing manufacturing in production manufacturing output per assets / sales manufacturing facility since 89 (in real terms or U.S. dollars) Centralize Concentration of Percent change 3 None 3 0.5 Fixed 0.005 0.00075 warehousing warehousing in shipments per warehousing warehousing assets I sales facility since 89 (in real terms or U.S. dollars) Ability to lower costs of the change in the change in (less N/A NO N/A N/A Percent N/A 0 ship transportation fraction of goods than truck load savings from products shipped without shipments)/(total switching away without hiring a full truck shipments) from from full truck hiring a full 89 to 97, in tons loads truck or dollar value. 38 RATIONALIZING LOADS Adopt standard Lower costs of Average size of a Percent change 3 Sending out 0 0.6 Ouibound 0.2 0 lot sizes handling and shiprnent to in average larger truc:king + other transportation customers (units outbound shipments trans. /total or tons per shipment size does not leai trucking + other shipment) since 89 to any cost trans. savings Containerization Decrease Damagecd goods Reduction in N/A NO 0.005 0 1 0 damages and and losses in (losses in losses in transit transit as a transit)/sales percentage of from 89 to 97 sales (annual) Allow cheaper Increase in N/A End of 0.05 1 Percent cost 0.3 0 transportation (containerized government savings from opportunities transport) / (total subsidized shipping in a transport) from container container 89 to 97 recycling inslead of other program mel:hods (open reduced the bulk or growth of packaged) containerizatiDn MODALITY Exploiting Lower Change in the Increase in non- -0.02 Better trucking 0 0.8 Percent cost 0.1 0 multi-model transportation composition of truck transport / services led to savings from transportation costs, faster transport use total a substitution using other opportunities service (trucks, rail, transportation away from rail modes instead maritimel from 89 to 97 of Irucking INVENTORIES Higher quality Using just-in-time Reduction in in- Reduction in 1.5 NO 1.5 0.7 Conversion of 0.003 0.000287 transport delivery of raw bound inventory average days of days 671 services allow material, parts, levels in-bound more timely and components inventory from deliveries or cuts down 89 to 97 shipments inventory levels Build-to-order Reduction in Chainge in out- Reduction in -30 Without the 60 0.7 Conversion of 0.003 0.011506 manufacturing finished goods bound inventory average days of trucking dalys 849 inventories levels out-bound improvemerts, inventory from the average 89 to 97 outbound inventory would be 4 months (instead of l). Reduc:tion in Obsolete Reduction in N/A N/A 1 1 0 obsolescence products written obsolescence/ off or marked sales from 89 to down 97 WAREHOUSING Restructure Lower costs of Warehousing Percent N/A 0 1 Number of N/A 0 and warehouse operations require reduction in wcrkers in computerize operation fewer workers (number of warehousing/ warehousing workers)/ total number of operations (annual volume employees of warehouse operations, in tons or pesos) from 89 to 97 Warehousing Percent N/A 0 1 warehousing N/A 0 operations require reduction in assets/fixed less capital (total assets warehousing assets)/(annual volume of warehouse operations, in tons or pesos) from 89 to 97 INSURANCE Get Limit losses Total transit loses = Reduction in 0.01 This reduction 0 0 1 0 comprehensive (theft & uninsured losses + total transit has nothinc to insurance damage) in insurance premiums loses/sales from do with coverage transit F loss prevention 89 to 97 insurance. costs Cheaper How much the Increase in 0 NO 0 0 (Loss claims 0 0 39 insurance company benefited (insured from insurance costs (lost) from cheaper freight)/(total companies)/(pre (more expensive) freight) from 89 miums paid) in insurance to 97 a typical year DOCUMENTATION Faster and Reduce losses Uncollectable Reduction in N/A Increased 0.005 0.3 1 0 reliable due to missed receivables = uncollectable workload of the documentation or untrackable shipments lost or receivables/ court system (enhanced deliveries unverifiable due to sales from 89 to makes it property rights) (goods deficient 97 harder to shipped documentation prosecute without reliable fraud proof of delivery) Enhanced More quickly Trade Reduction in N/A 1994 crises 2 0.8 Conversion of 0.003 0 negotiability receive receivables/sales average days forced us to days (liquidity) payment for receivable from extend longer goods sold 89 to 97 payment terms BOTTLENECKS Eliminate Faster Reduction in idle Reduction in N/A NO 0.03 0.2 (Total trucking + 0.06 0 bottlenecks clearance of time (idle time = average idle other transport formalities total time without time/total transit costs) / sales unnecessary movement) time from 89 to waiting 97 Faster Reduction in idle Reduction in N/A NO 0.03 0.3 (Total trucking + 0.06 0 loading/unloadi time (idle time = average idle other transport ng at ports, rail total time without time/total transit costs) / sales stations, movement) time from 89 to warehouses 97 Faster Reduction in idle Reduction in N/A NO 0.03 0.3 (Total trucking + 0.06 0 customs time (idle time = average idle other transport clearance total time without time/total transit costs) / sales movement) time from 89 to 97 TOTAL 9.7% Notes: 1. Changes that have taken place at the firm level since deregulation/ increased competition. 2. Specific areas that potentially could be affected by increased competition in trucking services. 40 Appendix 3. Four Case Studies 1. Fertilizer Coinpany A Mexican. fertilizer company producing and distributing urea and DAP, creatted in 1993, provides one of the starkest examples of the downstream benefits of rmore competitive, market- driven trucking services. In terms of inbound logistics, roughly 80 percent of the raw materials arrive by ocean shipment to the nearest port with the remainder arriving by train from the United States. In the end, 94 percent of the inbound raw materials make the final leg of the trip to the one production plant by truck. In terms of finished goods outbound logistcs, 98 percent is transported by truck to clients, with 70 percent of this total volume accounted for by trucking services directly contracted for by the company and the remainder picked UI) at the plant by various larger clients (retailers) who have their own in-house trucking fleet or outsource it. Inbound and outbound trucking costs jointly account for 19 percent of total costs. Total trucking costs account for 21 percent of sales (total transport costs, all modes, account fcr 32 percent of sales, of which 65 percent is trucking, 30 percent ocean shipping and 5 percent rail shipping). In terms of number of trucking service suppliers ideally suited to meet the company's inbound logistics reeds, there was an increase from 2 to 4; in terms of outbound logistics, the number increased from 6 to 9, confirming an incrceased intensity of competition in service supply. In turn, the intensity of competition at the downstream level has also increased, with the number of direct competitors that could take over a significant part of its operations increasing from 4 to 10. The cost faced by this company of hiring a truck on a standard route has reportedly fallen by 15 percent in real terms over the past five years. If the truckers were not to use toll roads, they would be able to offe:r the service for an additional estimated 6 percent less. The quality of the trucking services also increased, with losses in merchandise during transit falling by 50 percent. Of this reduction in total transit losses, 80 percent is estimated to be due to higher quality trucking services, with the remaining 20 percent attributed higher levels of security provided on federal roads. Higher quality transport services also allow more timely deliveries and shipments, with significant cost reductions from lower inventory holding costs. The average days of inbound inventory fell 87.5 percent, from 2 days to 6 hours, with 70 percent of this change reportedly due to improved trucking services and the remainder due to improved market conditions. The average days of outbound inventory rose from 1 to 2 months, but without trucking improvements the average outboumd inventory reportedly would be 4 months (so it is actually 50 percent lower than otherwise). These changes, in turn, have had a significant impact on innovat:ions and new products introduced by the company. Since the onset of regulatory reform, thc company has increased the number of distinct products that it sells from 12 to 32, with these chEnges reportedly almost totally due to improvements in trucking services. Sales of new products introduced over the past years represent 30 percent of total current sales. Of this increase, 70 percent is reportedly directly attributable to lower cost and higher quality trucking services, with the remaining 30 percent attributable to growvth in demand due to the higher level of fertilizer use in Mexican fields. The 41 average distance in trucking-based transit from plant to sales points has increased three-fold, from 60 to 180 kilometres, suggesting that many locations that previously were not profitable now have become so. Delivery times have also been reduced by 50 percent, with a regular trip that used to take 8 hours now taking 4 hours, reportedly mainly due to better equipment and a reduction in number of stops taken by drivers. Number of employees have increased from 20 to 60. Based on an intensive matrix survey instrument administered to this firm, the overall improvement in operating margin directly attributable to road freight innovations is 9.7 percent: reductions in trucking costs directly related to road freight innovations accounted for 4.4 percent, gains from new or improved products accounted for 3.5 percent (including reaching new areas and more direct deliveries), and savings from improved logistics accounted for the remaining 1.8 percent (including savings from centralized manufacturing and warehouses).62 2. Vegetable-producing Company Another type of company where trucking services account for a substantial part of total costs is an agroindustrial company that produces and distributes a specific class of Mexican vegetable, peppers. The company sells 10 percent of its production in the Mexican domestic market, exporting 90 percent with all foreign sales going to California (the company accounts for 4 percent of the entire California pepper market). Trucking costs account for 18 percent of total sales (the company does not use any other transport modes). Between 1991 and 1997, the total logistics cost of the company decreased from 42 to 23 percent of average price, with trucking being the logistics service with the greatest impact on costs. In real terms, trucking costs are reported to have fallen by 5 percent. An important additional cost element that has been falling substantially is the cost of inventory holdings. Input inventories have fallen from 45 to 18 days on average, or 60 percent. Inventories of finished goods have fallen from 7 to 3 days. Without doubt, an important driving force of these changes has been the increased intensity of competition in trucking services, with the number of providers that can offer the required specialised trucking delivery services of finished goods for this type of company increasing three-fold, from 5 to 15. Since 1991, sales in new areas as a fraction of total sales account for 90 percent of total sales. The respondent was not able to estimate, however, the extent to which this increase was due to cheaper or higher quality trucking versus other factors. Certainly, idiosyncratic factors such as the development of an array of personal contacts in California played a big part in the increased potential sales outlets. Improvements in the quality of trucking services, however, were estimated to be fally responsible for a 2 percent reduction in losses from transit-related problems (which fell from 5 percent of sales to 4 percent of now higher total sales). A fall of 10 percent in delivery times was also reported to be due to changes brought about by trucking deregulation. Number of employees have increased from 12 to 52. 62 See Appendix 2. 42 3. Eiectronie Components Company A large Mexican producer of electronic components provides an interesting picture of a company for which transport and logistics costs represent a much smaller percentage of total costs. The company has downsized, with total employees falling from 500 1:0 350 between 1989 and 1997. It produces mainrly passive components such as elecirolytic capacitors (50% of total current sales), polyester, polypropylene and ceramic capacitors, and resistors. They build t:o order from one production plant, and sell only to wholesalers. The electronic components are shipped to the United States (Michigan) along the Mexico City - Laredo corrido-, and to the airport for container shipments to Europe and Asia. For U.S. shipments, they use the services of a smaller trucking company who transports the product to the border. The prcduct is then transferred to a larger company for within-U.S. transport. However, the U.S. trucking company is responsible for the entire trip and quotes a single price. Total trucking costs account for roughly 2 percent of sales (total transport costs, aLir and trucking, are double, at approximately 4 percent of sales). In terms of number of trucking service suppliers ideally suited to meelt the company's outbound logistics needs, there was an increase from 5 to 10. The intensity of competition at the downstream level increased in each individual product category, on average from 2 otheir firms that could take away a significant part of shipments to 4. The cost of the Mexico City to Michigan route has increased by r oughly 15 percent in U.S. dollar terms between 1]993 and 1997. However, the respondent was not able to provide any estimates on the impor-tance of reversing factors, such as increases in the cost of toll roads, higher quality trucking services or more timely deliveries. It is clear that the quality of tmcking services increased substantially, with transit-related losses declining by a spectacular 90 percent. Half of this increase is reportedly due to trucking services, while the remainder is due to improved in-house packaging. Delivery times on a given segment have also decreased by 40 percent. The main impact of improvements in trucking services appears to have been in expanding sales, with sales in new areas accounting for 50 percent of total sales, and 30 percent of this increase accounted for by higher quality trucking services. In addition, there has been a 20 percent increase in direct delivery customers, and in this case 70 percent of the increase is reported due to higher quality trucking services. There also has been a significant increase in new products, with the range of individual products in stock rising from 100 to 200; in terms of value of sales, new products only account for a reported 7 percent of total 1997 sales. In this case, improved trucking services accotut for less than 20 percent of this increase, with most of the change attributed to the increased price attractiveness of individual components as a result of Nafta-relateid tax changes and to internal measures to raise quality. There have been a number of olher logistics-related internal changes, such as introducing bar code tags to improve maeerial handling and introducing computerised systems and telecoms improvements for faster and more reliable documentation handling. 43 4. Electronics Company A large multinational electronics company provides a useful illustration of impact of road freight reform where the cost of logistics is a very small fraction of total costs. The company employs 1330 workers, a slight downsizing from 1500 in 1989. It mainly produces semiconductor- rectifiers (accounting for 40 percent of sales) and 'surmeticos' (for an additional 20 percent), in addition to an array of other electronic components for export. Most of the output (80 percent) is for in-house use in other worldwide production and assembly plants, with the remainder sold directly to other companies. Total trucking costs account for 1.8 percent of sales (total transport costs, all modes, account for 4 percent of sales, of which air cargo is 50%, trucking 45% and ocean shipping 5%). In terms of inbound logistics, roughly 60 percent comes from the United States via truck, and the remainder from Europe and Asia via air and sea. In terms of outbound logistics, which is more time-sensitive, roughly 30 percent is shipped back to the United States via truck while the remainder is shipped by air to Europe and Asia. Of the five more intensively surveyed companies, it is the two electronics component producers for which air cargo is the logistics service perceived to have the greatest impact on costs. In terms of number of suppliers that are ideally suited to meet the company's inbound and outbound trucking requirements, the number has increased from 1 to 2. In terms of intensity of competition as measured by direct competitors that can take away a significant part of shipments, whereas the company was the only game in town in 1989, there are now 2 other close competitors. The cost of a standard trucking route to end-points in the United States has increased in real terms by roughly 5 percent. The respondent mentioned a number of reversing factors, including newer trucks and tracking systems, but could not accurately estimate the impact of those improvements. The main improvements appear related to quality of service, with losses related to logistics problems having decreased by 75 percent (from 4% to 1% of sales). In addition, delivery times on a given segment have decreased by 20 percent. However, it was again not clear to the respondent how much of the increase was actually due to improved road freight services versus other factors such as improved loading and unloading facilities and more sophisticated internal software that facilitates planning and tracking. In comparison with 1989, the firm has introduced new products, with the stock of individua.l products increasing from 20 to 30. Sales of new products account for 15 percent of total 1997 sales. The firm also covers a wider geographic area, with sales in new areas also accounting for roughly 15 percent of total 1997 sales. Finally, there has also been a significant increase in the number of new direct-delivery customers. However, in all these areas, it was difficult for the respondent to quantify with any degree of certainty the proportion of such new sales attributed to improvements in road freight services. It is possible that part of the difficulty especially in this instance stems from the inherent additional complexities of a multinational firm where key strategic sourcing and logistics decisions are made in headquarters located in another country. 44 FIGURE 1: Road freight haulage units of the Mexican public federal system, 1980-1996 350,000- 300,000 | Motor units C Trailers 250,000- 200,000- 150,000 Tri iiI 100,0~0 50,0(00 o - .0 ' 00 Go Go Go o a, OV I o. j 0.,,,, ON O N O ~O N O , 0, ,,,- O O tOr.- Mottr Year ui its Trailers Total Year unit s Trailers Total 198Tr7-0,1 31,20 141793W 1r989 124,87- 39,11~73 W14,1 1981 117,090 36,760 153,850 1990 178,130 44,853 222,983 1982 119,309 38,216 157,525 1991 209,060 67,865 276,925 1983 120,702 38,639 159,341 1992 224,913 78,233 303,146 198 4 120,702 38,639 159,341 1993 232,203 81,307 313,510 198.5 117,956 38,317 156,273 1994 198,273 93,827 292,100 1985 114,000 36,500 150,500 1995 204,117 96,638 300,755 1987 115,092 37,245 152,337 1996 212,909 102,409 315,318 1988 115,897 37,506 153,403 Source: Based on SCT (1996), Estadistica Basica del Autotrans3orte Federal 45 FIGURE 2: Road freight enterprises according to number of haulage units, 1996 60,000 160,000 50,000Q. =No. ofenterprises + No. of hauluit 140,000 50,000 000S:f 2C0 120,000 40,000 100,000 30,000 80,000 60,000 10,000 20,000 0 0 C~~~~~~~~~~~~~ Type of enterprise No. of units per No. of No. of haulage uiits enterprise enterprises Owner-operator 1 - 5 54,292 144,158 Srnml 6 -30 4,991 90,052 Medium 31-100 1 53,032 Large 101 + 187 28,076 Total , 315,318 Source: Elaboratedby CANACARbased on information fromINEGI, Censos Xl y Xfl de Comunicaciones y Transportes. 46 ]FIGURE 3: Quality-price trade off ir the road freight industry Price Mexico - after U.S. ' Mexico - before Quality 47 FIGURE 4: Age of equipment (motor units) 1990-97 and projections 1998 -2010 : New fleet purchases M Renovations ---Ave rage ap e n y e 40,000 a a 20,0001.I -4 I 000; z,