Maquila, Economic Reform and Corporate Strategies

Transcription

1 World Development Vol. 28, No. 9, pp. 1627±1642, 2000 Ó 2000 Elsevier Science Ltd. All rights reserved Printed in Great Britain X/00/$ - see front matter PII: S X(00) Maquila, Economic Reform and Corporate Strategies RUDOLF M. BUITELAAR Economic Commission for Latin America and the Caribbean, Santiago, Chile and RAM ON PADILLA P EREZ * Economic Commission for Latin America and the Caribbean, Mexico Summary. Ð Corporate strategies in maquila have changed over time. While they were originally geared only to the exploitation of low-paid, low-skilled labor, nowadays rms are upgrading human capital and increasing the scope and technological complexity of production processes. The adoption of the New Economic Model in host countries aimed precisely at this type of strategy change, but cannot solely be credited for it. In addition, policies regarding maquila under the NEM did not achieve all the stated goals regarding corporate strategy change: trends in local procurement and technology transfer are still unsatisfactory. Strengthening of local supply and absorption capabilities is required. Ó 2000 Elsevier Science Ltd. All rights reserved. Key words Ð Mexico, Central America, maquiladoras, employment, technology transfer 1. INTRODUCTION 1627 The evolution of maquila o ers an interesting case for the study of the impact of economic reform on corporate strategies. Maquila has evolved into a major employer and key export sector in Mexico and in a number of Central American and Caribbean (CAC) countries. This paper concentrates on the Mexican case but also includes a comparison with the CAC maquila. The maquiladora program is an investment attraction and export promotion scheme that o ers bene ts to quali ed rms regarding import duties and other taxes (CEPAL, 1998a). The program was initiated in Mexico in the mid-1960s to ameliorate the high unemployment rates in the northern border zone that had resulted from the termination of the braceros program. It consisted of a facility to import duty-free raw material, intermediate inputs, and capital goods from the United States, for the production of goods to be re-exported to the United States. But the facility was not part of Mexican import-substitution industrial policy. It had some quite restrictive clauses in order to prevent maquila production from competing with local industry. As a consequence, rms from the United States and Mexico limited their use of the program to exploiting the immediate advantages of lowskilled and low-paid Mexican labor force. They used the bene ts of this program and the US tari item to have a few simple assembly tasks carried out in Mexico. Similarly, maquila industry was established in the CAC region in the 1970s, with the creation of Export Processing Zones (EPZs) whose main goal was to create employment for the regionõs low-skilled labor force. EPZs were not very successful, however, until the 1980s when economic and political reforms were adopted and tari bene ts on the US market were obtained under the Caribbean Basin Initiative (CBI). * The authors wish to acknowledge the contributions of Ruth Urrutia to the eldwork and analysis of results in general. Andres and Elizabeth Dauhajre organized the Dominican Republic interviews. Ronney Zamora of the Costa Rican Chamber of Industry and Brenie Matute of FIDE in Honduras were extremely helpful in organizing eldwork in their countries. Wilson Peres, Nola Reinhardt and Jorge Mattar provided useful comments and guidance.

2 1628 WORLD DEVELOPMENT Under the New Economic Model (NEM), starting with the economic reforms of the mid- 1980s, the Mexican Government and the participating rms began to see maquila as an integrated part of export-led development and its speci c legal framework was transformed. The aim was to change the corporate strategies, promote investment in human capital and the transfer of technology, and foster integration with local Mexican manufacturing industry. In the CAC region, a similar reorientation under the NEM has taken place, to a limited extent, in Costa Rica. Critics of export-led development still refer to maquila to illustrate that opening-up only means low wages, disarticulation of production chains, few backward and forward linkages, no technological e ort or learning processes, and subordination to the strategies of foreign rms. Thus, two central questions arise that are the subject of this paper. First, to what extent have maquila rms indeed changed their competitive strategies from exploiting only the static gain of low wages to creating dynamic gains through training, technological learning and local networking? Second, what role has economic reform in Mexico played in bringing about any change in corporate strategies? The paper is organized as follows. First, a brief overview will be presented of the origin and evolution of maquila in Mexico, against the backdrop of Mexican and US public policies, international competition and technological change. Second, the performance of the maquila program and its corporate strategies with regard to its stated objectives are analyzed. This section uses the results and insights from a rm survey conducted by the authors in 1997 (CEPAL, 1998a). 1 The survey was also applied to maquila rms in Central America and the Dominican Republic, 2 and the third section compares the situation in these countries with that of Mexico. The paper concludes with a short re ection on regime change and corporate strategies. It has to be pointed out that maquila is not an industry in the usual sector-speci c sense of the word. The maquila programs and laws constitute an investment attraction and export promotion scheme. Hence, a great variety of maquila rms exist in di erent industries. Two important common features have been suggested in the literature: the orientation to the market of the United States, and the intensive use of relatively low-cost labor (Carrillo, 1997). 2. ORIGIN AND EVOLUTION OF MAQUILA FOR EXPORT IN MEXICO The maquila program is a result of the intricate social, economic and political relations between Mexico and the United States, in the context of increasing global competition and rapid technological change. To start with the latter two points, maquila should be understood as a phenomenon pertaining to the international reorganization of production in the second half of the 20th century, as ttingly described in the product cycle theory. According to Vernon, the introduction of a new product on the market requires highly quali ed personnel performing research and development activities and innovative production processes. Hence, these tasks are located in industrialized countries with a comparative advantage in this type of worker. When technology matures and the production process becomes standardized, the skill intensity of production diminishes. In consequence, the international competition will result in the transfer of the production of mature goods to developing countries with a comparative advantage in relatively less skilled workers (Vernon, 1966). From the 1950s on, Japanese rms started to pose a challenge to US manufacturing industries. The relocation of labor-intensive production processes from Japan to other Asian countries in the 1960s and 1970s enhanced the challenge. In response, US manufacturing rms looked at Mexico and the Caribbean Basin as its reservoir of less skilled labor to counter the Asian challenge and move south its labor-intensive production processes (Frobel, Heinrichs & Kreye, 1980; Grundwald & Flamm, 1985). The US government incorporated this idea in its trade policy, designed to strengthen the competitiveness of US rms. Tari item , introduced in 1963 following the earlier example of tari item , is a clear expression of this strategy. Nowadays, these items are included in the Harmonized System as and , respectively. The latter is the most important, and permits the importation of goods assembled in a foreign country with components manufactured in the United States. The duty is calculated only on the value added abroad. The duty-free imports under the maquila program in Mexico came to complement nicely the tari provision, as did the same facilities in Export Processing Free Zones

3 MAQUILA, ECONOMIC REFORM AND CORPORATE STRATEGIES 1629 (EPFZÕs) in Central America and the Caribbean. Thus, inputs and raw material from US origin could cross the border and the nal good could be re-imported into the United States without much red tape or high duties. The policy to enhance US competitiveness by using cheap labor south of the border was supported by additional measures. Important tari concessions were granted by the Generalized System of Preferences (1976), which allows the duty free entry of some products that come from selected developing countries (CEPAL, 1996), the Caribbean Basin Initiative (1984) and MexicoÕs Special Access Program (1989). As is well known, the nature of the Asian challenge, or global competition for that matter, has changed signi cantly. In the 1960s, low production costs were central. Later the price-quality relation came to the fore. The introduction of quality as a central feature of competitiveness made certain production technologies and organizational systems obsolete. The so-called Fordist production model, based on TaylorÕs management theory, gave way in the 1980s to post-fordist production techniques. Quality control and teamwork were introduced, which increased the skill requirements. Increasingly, for US rms to win in global competition, it was not enough just to use cheap labor from Mexico and the Caribbean Basin for certain labor-intensive production processes. US policy, at least regarding Mexico, became oriented to full commercial integration, culminating in the signing of the North American Free Trade Agreement. Firm strategy changed from transferring labor-intensive production processes to Mexico, to a complete reconsideration of their production and marketing strategies from a truly North American point of view. Global competition and technological change, and the corresponding strategies of the US government and rms, are important to understand the origin and evolution of maquila, but present only one side of the story. From a more Mexican viewpoint, the speci c history of the maquila program started in In that year the US Congress, under pressure from labor unions, decided not to renew the bilateral agreement with Mexico known as the Programa de Braceros, under which Mexicans were allowed temporary work permits in the United States. As a result, 200,000 Mexican workers became jobless, and the rate of unemployment in border towns reached levels between 40% and 50% (Urias, 1978). In response, the Border Industrialization Program (Programa de Industrializacion Fronteriza, PIF) was established in Its main objective was job creation, through the attraction of foreign direct investment to establish assembly operations for export in the border zone. Additional goals were to upgrade skills, to stimulate demand for local supplies, and to lower the trade de cit. The PIF became operational in It applied only to zones within 20 kilometers of the border, allowed duty-free imports of raw material and capital goods, obliged rms to re-export all of its production and required at least 51% national capital (Urias, 1978). Unemployment was a problem not only in the border zone, and the PIF program also seemed a feasible solution elsewhere. In 1971, the Mexican Government extended the program to coastal areas, and in 1972 to the whole national territory, except industrialized areas. Moreover, in 1973, full foreign ownership was allowed. From 1977 onward, the maquiladora program was even used to help domestic rms producing for the local market, in low points of the business cycle. Firms with idle capacity could qualify to bene t from the maquiladora program, provided at least 20% of their inputs and raw material were of local origin (CEPAL, 1996). Until the beginning of the 1980s, the maquila program was an exception to the import-substitution industrialization regime. The overall economic and political environment was not particularly receptive, except for the speci c border towns where the rms operated. The regulatory framework and the general attitude toward foreign capital were an implicit strain on the growth of the industry. The laws for technology transfer registry and for foreign investment, issued in 1972 and 1973, respectively, gave the government the power to review and control every technology transfer contract. The installation of foreign rms that could harm or compete with national industry could be prohibited (Gonzalez, 1990; Mexico, Government of, 1972, 1973). In the 1980s, fostered by the debt crisis, a process started that would result in the gradual elimination of the special status of the maquiladora rms, and a more complete integration of Mexico in the North American production structure. Under the NEM, market mechanisms and the opening of the Mexican economy to international trade became the central policy

4 1630 WORLD DEVELOPMENT guidelines. The maquila program became one of the most important mechanisms of export promotion (De la O, 1997). 3 In 1983, maquiladora rms were allowed to sell up to 20% of their production on the domestic market (Mexico, Government of, 1983). In 1989, new legislation to regulate and to promote the maquila program was adopted (Mexico, Government of, 1989). Two new goals of the program were added: (a) to contribute to sectoral integration and increase the competitiveness of domestic industries, and (b) to stimulate the development and transfer of technology in the country. The ability to sell on the domestic market was enhanced (up to 50% of exports or one third of total production; Mexico, Government of, 1989). This legislation also established the goal of an annual growth superior to 10% in employment and foreign exchange generation. A series of regulations were eliminated to make the legal framework more transparent and to have easier and faster scal and administrative procedures (Sanchez, 1990). Meanwhile, the process of opening up the Mexican economy was implemented rapidly: during 1985±90 the import permits were reduced from 92% to 20% of tari items, and the average import tari was cut from 24% to 13% for the same period (Mendiola, 1998). Of outstanding importance in the NEM were also the new regulations to create a clearer economic and administrative environment. Of particular importance for the maquila sector were the new laws for foreign investment and technology transfer, which reduced uncertainty and allowed longer-term planning of operations. In 1993, with the NAFTA already in sight, a new regulation established the nal agenda of market fusion: in 2001, after a schedule of progressive integration, maquiladora rms will be allowed to sell all their production on the domestic market. Finally, aiming at greater integration with local industry, in 1994 the maquiladora program was extended to rms supplying goods and services to maquiladora activity (Mexico, Government of, 1994). The most recent regulation was issued in June 1998, with the main purpose of simplifying the administrative processes (Mexico, Government of, 1998). In 2010, NAFTA will have provided full duty-free access of products of Mexican origin on the US market (Mexico, Government of, 1993). At this point, the maquila program will cease to be operational in the US±Mexico commercial relation (but not for exports outside NAFTA). This does not mean that assembly operations for export will come to a halt in Mexico. According to the Secretary of Commerce and Industrial Development (SECOFI), these ``will continue to be one of the most attractive forms to participate in the processes of globalization and production sharing.'' Moreover, the agreements reached under NAFTA regarding apparel and vehicles will bene t the assembly activity (SECOFI, 1993). If anything, maquila plants will start to cater also to the domestic Mexican market and local rms will seek to strengthen their position on the US market by incorporating assembly operations. It will become increasingly hard to tell the di erence between assembly and manufacture and both will be used for foreign and domestic markets alike. TodayÕs maquiladora rms are very likely the forerunners of the integrated North American rms of the early 21st century (see Table 1). 3. MAQUILA: GLOBAL PERFORMANCE AND CORPORATE STRATEGY In this section, the evolution of the maquila industry and its corporate strategies are analyzed, following the goals set out for the program in the course of its history. (a) Employment generation From the start, the rst and foremost objective of the maquila program was job creation. In the rst 15 years the evolution of employment in maquila industries was closely linked to the evolution of the US economy. The recession in the US of 1973±74 resulted in an important decline of employment in maquila. Over time, the link became less clear and the recession of 1981±82 produced little employment loss. The 1990±91 recession meant merely a slowdown of the growth rate (CEPAL, 1996). During 1982±86, a process of consolidation of the maquila for export industries may be observed. In the midst of a severe crisis in Mexico, the activity proved to be most solid in terms of the creation of rms and jobs (CEPAL, 1992). Moreover, 80% of the maquila employment has been created since the economic reform (mid-1980s), and over 400,000 jobs were created in NAFTA period alone (1994±98).

5 MAQUILA, ECONOMIC REFORM AND CORPORATE STRATEGIES 1631 Table 1. Mexico: economic performance (1970±98) a 1970± ± ± ± ±98 GDP (annual average growth rate) ) Manufacturing industry (average ) annual growth rate) Domestic gross investment/gdp (period average) Foreign direct investment (period ,078 4,683 10,396 average) Consumer prices index (average annual rate) a Source: Banco de Mexico, INEGI, CEPAL, Mexico-O ce Data base. As can be observed in, Table 2, maquila employment has increased without interruption since In 1998, 1.04 million direct jobs in 3,130 plants were reported. This represents 42% of manufacturing employment and 9% of total employment. Also important, but not measured, is indirect employment. Estimates vary: Guajardo estimates indirect jobs in local supplier rms at around 44% of direct employment, and jobs generated because of the multiplier e ect of maquiladora wages at around 73% of direct employment (goods and services purchased by maquiladora workers) (Guajardo, 1992). On the other hand, Carrillo estimates that 80% of maquila employment are direct jobs (Carrillo, 1997). In any case, maquila has proven to be a very successful device for employment generation, principally since the mid-1980s. In fact, employment in maquila has grown at around 15% per year during 1994±98, whereas employment in manufacturing industry in Table 2. Mexico: evolution of maquiladora industry for export (1974±98) a Year Firms Workers Gross production value b Imported inputs Value added Local inputs Wages (as % of gross production value) ,974 10, N.A ,214 12, ,496 17, ,433 25, ,704 35, ,365 50, ,879 57, ,102 78, , , , , , , ,544 1,305, ,388 3,445, , ,743 9,823, , ,245 22,846, , ,064 30,676, , ,474 39,578, , ,146 47,742, , ,035 58,584, , ,588 72,082, , ,585 90,100, , , ,237, , , ,692, , , ,230, ,130 1,038, ,050, a Source: Instituto Nacional de Estadõstica, Geogra a e Informatica (INEGI), b Thousands of current Mexican pesos.

6 1632 WORLD DEVELOPMENT general (maquila not included) has not grown signi cantly since In 1998, total direct employment in manufacturing industry (without maquila) hovers around 1.45 million. 4 If present tendencies continue to hold, in ve years maquila employment will outnumber employment in other manufacturing industries together. Two related tendencies are observable. First, job growth is faster in inland areas than in the border zone. In 1988, 7% of maquila jobs were outside the border zone; in 1998 this percentage has increased to 21%. But Tijuana and Ciudad Juarez continue to be the two most important cities regarding maquila employment. The second tendency is that apparel industries register faster job growth than other industries. In 1988 close to 9% of all maquila jobs were in apparel industries; in 1998 this percentage has increased to 21%. Apparel industries now generate approximately as much employment as the car parts subsector (19%) and electronics (34%). The two tendencies combine in the fact that fastest growth is recorded in apparel industries in states such as Oaxaca, Tlaxcala, Puebla and Yucatan. The surprising growth of apparel maquila may be explained principally by NAFTA. As a consequence of this agreement, 45% of textile and clothing exports from Mexico to the United States were exempted immediately from import tari s, while the rest of products will be exempted gradually before the year Furthermore, NAFTA eliminated 90% of import quotas applicable to this sector (CEPAL, 1995). Average plant size has increased from 261 to 332 workers per plant during 1988±98. From the start of NAFTA, the plants increased their size by 14.2%, which is a third of the overall increase over 1982±97. For example, Lucent Technologies and Attel del Norte, recently acquired by Phillips Electronics, were motivated by economic deregulation to open new plants and expand their operations. As a matter of fact, thanks to NAFTA, a new program was started to sell locally its products on the telecommunications market. In the same sense, over the past three years in Guadalajara US$850 million have been invested by rms such as Kodak, IBM and Hewlett±Packard (Mendiola, 1998). Regarding wages, Table 3 shows that average wages in maquila industries are around 62% of average wages in manufacturing industries. Di erences in skill composition of labor force, Table 3. Average wages in manufacturing industry and in maquila industry (current mexican pesos per worker, daily rate) a Manufacturing industry Maquiladora industry a Source: Secretarõa del Trabajo y Prevision Social, and also di erences in unionization, may account for the gap. 5 There appears to be a trend break around Whereas during 1977±92 real wages fell 45% in maquila industries against a drop of 30% of real wages in manufacturing industry, during 1993±98 real wages in maquila industries performed better than those in manufacturing industries: an 8.1% increase versus a 10.3% decrease (INEGI). The trend of increasing demand for skills in maquila industries may account for this. For skilled labor, maquila wages compete with nonmaquila wages (Gambrill, 1995). Administrative personnel and technical employees have increased their participation in the maquila labor force: the former increased from 5.8% to 7.2% during 1980±98, and the latter from 9.2% to 12.1 in the same period. 6 (b) Human resource development 7 The maquila program was always considered to be a possible vehicle for the upgrading of skills and the formation of human capital. The rst generation of maquila industries demanded only limited skills, however, o ered few training facilities and established little relations with local educational institutes. Over time, with the introduction of post-fordist production techniques as a response to more intense competition and a demand for higher quality products and with the evolution of maquila tasks in terms of complexity and integration, the quality of human resources became a more important competitive factor. The ECLAC-Mexico survey con rmed that rms started to hire personnel with more formal education. In addition, the interviewed rms o er in-house training courses, most of them to introduce workers to the rm system and teach job-speci c skills. General training

7 MAQUILA, ECONOMIC REFORM AND CORPORATE STRATEGIES 1633 courses are o ered on average twice per year, on quality control, cost reduction, security, etc. General interest courses are o ered on social skills, prevention of sexual harassment, personal hygiene, and so forth. For higher-level employees, ``train the trainer'' courses, customer satisfaction, total quality control and speci c technical skills are o ered at least once a year. For administrative personnel, once or twice a year, courses are o ered on teamwork, scal and labor issues, and customer satisfaction. 8 These courses are taught by quali ed own personnel or by experts from a great variety of educational centers. Consequently, more than 50% of rms interviewed reported to have formal agreements with educational institutions. The most common one is to accept internships that give the student the opportunity to get acquainted with a rm and the rm to select the professionals it needs. Occasionally, an employee from a maquila plant may teach at one of those educational centers. 9 Furthermore, skills are upgraded through an intensive process of ``on-the-job'' training of the learning-by-doing type. This derives from the assembly of products (the manufacturing process) and the application of the organizational system. The importance of this way of improving national skills depends on the complexity and technological sophistication of the production process, and the speci c responsibilities of the local work force. An important problem for maquila rms that has its consequences for the contribution of the rm to the improvement of human resources, is the high rotation of the workforce. In the interviewed rms, average monthly rotation is 4.9%. Some rms in Tijuana reported to the ECLAC-Mexico survey above 10% monthly rotation, whereas some in the interior of the Republic reported less than 1%. 10 This problem has induced the rms to implement a selective training strategy, concentrating training opportunities at the level of higher skilled workers. As a result, the training that less skilled workers may receive, at least initially, is quite limited. On the development of skills and learning among shop- oor workers, diverse opinions have been recorded during the interview. Some managers stated that there was no training for workers whatsoever, while others stressed the potential of maquila as a generator of knowledge and work experience. Most maquila workers are young and come from rural areas, and the maquila experience is their rst contact with the formal economy. In this sense, organizational culture and labor discipline are the main aspects of learning. Regarding technical and productive skills, learning among bluecollar workers is minimal. Their activities are very segmented and in most cases limited to the assembly of components. 11 In conclusion, maquila has contributed, albeit only to a certain extent, to the upgrading of the skills of the Mexican workforce. Firms have invested to some extent in training and the establishments of links with local educational institutions. A well-known theme surfaces also in the maquila activity: investment in education and training by the private sector is less than socially optimal because of the externalities involved. (c) Export promotion and the balance of trade Net foreign exchange generation as an objective of the maquila program used to be of secondary importance. Maquila was rightly assumed to be a very import-intensive activity. During the 1990s, however, activities under the maquila program became one of the countriesõ prime foreign exchange earners. Maquila exports report strong growth rates since 1983, similar to the employment growth rates. 12 Total exports grew at an average annual rate of 11%, whereas maquila exports grew at an average annual rate of 18.4% during 1980±98. Consequently, maquila exports as a percentage of total exports rose from 14% in 1980 to 45% in Likewise, maquila imports report an average annual growth rate of 19.4% against 10.3% growth rate of total imports. The share of maquila imports in total imports grew from 8.3% in 1980 to 34.5% in 1998 (see Table 4). As can be observed in Table 4, the trade surplus grew constantly, except for 1994±95. Data for 1998 show a trade surplus for maquila superior to 10 billion dollars. Since the overall trade balance shows a de cit of US$5.7 billion, it has to be said that maquila trade constitutes an important counterweight for the otherwise de citary Mexican trade in goods and services. Further comments on national content of exports are made in section (e). (d) Transfer of technology Technology transfer is ``the process by which knowledge related to the transformation of

8 1634 WORLD DEVELOPMENT Date Maquila exports Table 4. Mexico: foreign trade (1980±98) (thousands of dollars) a Maquila X/Total X Growth X maquila Maquila imports Maquila M/Total M Growth M maquila Total trade balance Maquila trade balance ,519, ± 1,747, ± )3,058, , ,205, ,229, )3,876, , ,825, )11.8 1,974, )11.4 7,044, , ,641, ,822, ,104, , ,904, ,748, ,184,153 1,155, ,093, ,826, ,398,196 1,267, ,645, ,351, ,019,703 1,294, ,105, ,506, ,787,089 1,598, ,145, ,808, ,609,529 2,337, ,328, ,328, ,054 3,000, ,872, ,321, )882,324 3,551, ,833, ,782, )7,279,040 4,050, ,680, ,936, )15,933,727 4,743, ,853, ,442, )13,480,576 5,410, ,269, ,466, )18,463,683 5,803, ,103, ,178, ,088,485 4,924, ,920, ,504, ,530,967 6,415, ,165, ,332, ,590 8,833, ,863, ,556, )5,742,180 10,306,977 a Source: Banco de Mexico, inputs into competitive products is acquired by national entities and whose source is foreign'' (UNIDO, 1995). This section analyzes changes in corporate strategy, the sophistication of the production processes and the degree to which local personnel is involved in these learning processes. 13 The transfer of technology as an objective of the maquila program was introduced only after the reforms of the 1980s. In the literature concerning this sector in Mexico, it is common to nd a distinction between rst and secondgeneration maquiladora rms (Carrillo & Hualde, 1997; Carillo & De la O, 1992; Wilson, 1990). The di erence is observable in organizational systems, in the importance of quality, in the application of di erent generations of production technologies and in human resource management. Recently, changes have been reported in the sense of a growing involvement of some maquila rms in design and some research and development. In some cases rms have taken responsibility for the complete production process. These new characteristics have inspired some researchers to use the concept of ``third generation maquila rms.'' Notwithstanding the evolution of maquila over time, it is important to point out that generations coexist (Mertens and Palomares, 1988; Brown and Dominguez, 1989). In fact, within rms it is possible to observe a technological duality, where traditional assembly activities are carried out alongside more capital-intensive ones using sophisticated technologies (Barajas, 1990; Rodriguez, 1990). The rst generation were virtually all rms of US origin doing traditional assembly operations based on the intensi cation of manual work requiring few skills. These plants were typically more interested in the volume of output than in its quality, as was usual in Fordist production processes. This type of plant was predominant from the start of the maquila program until the beginning of the 1980s. The second generation of maquila plants has its origin in the crisis of Fordism and in the new requirements of demand in the world market, in particular the need to produce higher quality goods and adapt swiftly to changes in demand. In this generation, which prevails since the beginning of the 1980s, rms of Asian origin start to participate in maquila. Buyer±supplier relations with local rms remain at a very low level, but the production process becomes less oriented toward assembly and more toward manufacture. Usually a higher technology level can be observed, with numerically controlled machinery and robotized production processes. Moreover, the level of autonomy of the maquila plants vis- a-vis the corporate headquarters or the main clients increases. They maintain a

9 MAQUILA, ECONOMIC REFORM AND CORPORATE STRATEGIES 1635 stable group of trained and skilled workers, have an increasing proportion of technicians and engineers, and have as a central source of competitiveness the improvement of quality. Regarding organization of production and work, important progress can be observed such as the introduction of just-in-time and total quality management, among other systems. It should not be forgotten, however, that the introduction of more sophisticated technology is aimed at obtaining more control, precision, homogenity and quality in labor-intensive production processes that remain the raisondõetre of maquila. Third generation maquila plants are distinguished by the performance of some design and research and development activities. Transnational corporations that use third-generation maquila plants develop intra rm networks. The technological dependency vis- a-vis corporate headquarters tends to disappear and a process of autonomous decision making may be observed. Highly quali ed personnel are the major part of the workforce. But, the source of competitiveness remains the relatively low labor costs (even of highly quali ed personnel) and the reduction in the duration of projects (Carrillo & Hualde, 1997). Also in the ECLAC-Mexico survey, a process of evolution of tasks has been observed that may be described as a trend toward more control over the production process and at the same time a sophistication of production technologies. Usually a maquila plant starts with a small assembly line, and upon satisfactory performance in terms of quality and e ciency, a greater number of tasks is transferred and a bigger part of the end-product is assembled and manufactured at the plant. This trend has to do both with the learning process at the plant itself, and with the economies involved in having complete production processes at one location. Thus, it is quite common to nd capital-intensive production processes and labor-intensive assembly lines existing simultaneously in one plant. For example, Mitsubishi recently expanded its circuits production in Mexicali, transferring the total production of television cards from Asia. The Taiwanese enterprise Delta Products is constructing a plant to assemble computer batteries which were produced in China and Thailand (Mendiola, 1998). It may be true that technological complexity has increased, but from the viewpoint of technology transfer the important question is to what extent local personnel and local management are involved in the learning process. From the ECLAC-Mexico survey, it becomes clear that local managers rarely participate in decisions of which end products to manufacture, or the selection of raw materials or inputs. Strategies concerning investment or nance, and production technologies, may or may not be in uenced by local management, depending on its relationship with corporate headquarters. Decisions regarding recruitment, human resource management and the organization of the production process are usually taken locally. Another study concludes that local management has increased its sphere of in uence and intervention, as a consequence of the process of delegation of authority to maquila plants and the displacement of expatriate management (Contreras, Estrada, & Kenney, 1997). Human resource management and bookkeeping were found to be among the rst tasks entrusted to local management, since these are indispensable to establish links with the local community. The most important learning processes and transfer of know-how are thus in the eld of organization. Among maquila plants it is common to observe the application of modern management systems. Total quality control; just-in-time; operations manuals, written procedures and internal norms and standards; programs of worker suggestions, are among the most commonly applied. The reasons to apply modern management systems are the high quality demands in the market place, the need for quick response to changes in demand, large volume of work in process, and the quest for lowest costs. 14 Regarding product and process technology, the learning processes are quite modest. In a few of the interviewed rms, a signi cant learning process has been observed among engineers and technicians who are in continuous contact with the complete production process. In most interviewed rms, however, process speci cations come from abroad, which means that learning is the product of adaptations and marginal improvements. Design and construction of production lines is the responsibility of foreign personnel, with some collaboration of local personnel. The rms in the ECLAC-Mexico sample report only very limited relations with laboratories and research and development centers, even less so when they are subsidiaries of a transnational corporation. As has been pointed

10 1636 WORLD DEVELOPMENT out, the technological policy followed by the transnational companies in maquila is to use turnkey projects. In consequence, the tasks related to the technological core are reserved for the research and development centers located in developed countries. (e) Interindustrial integration and the strengthening of domestic manufacturing through local procurement In the early years, there was mention of ``stimulating domestic demand'' as an objective of the maquila program. In later programs, after the economic reforms of the 1980s, maquila rms were supposed to ``strengthen the competitiveness of local industry'' and new regulations were enacted to foster local industry linkage to maquiladora activity. Either way, the goal is rather elusive. Maquila plants use notoriously little domestic input. Table 2 shows the evolution of the share of local inputs in gross production value of maquila rms. The historic low was reached in 1985, with local inputs contributing 0.7% to the gross production value of the industry. Thereafter the trend is upward, but the 2.2% reached in 1998 is still extremely low indeed. On the other hand, the national component of value-added shows a declining trend. Over a third (36.3%) of value-added was of Mexican origin in This gure declined steadily until reaching 18.6% in 1996, although it then increased in the last two years to reach 21.7% in This decline is due to the relative and absolute drop in the wage bill, which went from 22.4% of value-added in 1974 to 10.6% of value-added in Not only the drop in real wages, but also the increasing capital intensity of maquila production accounts for this trend. The above data explain why the total valueadded in maquila in 1998 was only about 12.7% of value added in manufacturing industries, while employment in maquila accounts for roughly one-half of employment in manufacturing industries. This is also the foundation for observations such as in Guajardo, that maquila has a low product multiplier factor compared with other industries. He estimated the product multiplier of apparel maquiladora plants at 1.58, against 2.52 in domestic apparel production. With estimates of multipliers for all maquila production, Guajardo calculates for 1989 the total production value of maquila at US$18.6 billion. Two-thirds of this value, or US$12.5 billion, corresponds to the direct production value. Another US$1.3 billion (7%) corresponds to the product generated in other sectors through local procurement of maquila plants, and US$4.7 billion (26%) is induced in other sectors of the economy as a result of expenditures of workers (Guajardo, 1992). It is thus clear that maquila plants are scarcely linked to the domestic economy. This is due both to factors within the plants and the maquila program itself and to supply constraints within domestic plants. Among the rst set of factors, occasionally the production structure in maquila plants is vertically integrated within a transnational corporation, and therefore local procurement may not be part of the corporate strategy. In addition, it has to be said that during the rst 20 years of the maquila program, the regulatory environment did not stimulate the development of a local supplier network. For one, the regulation of technology transfer was strict and not very transparent up to Strong controls, arbitrariness and few protection mechanisms against abuse of transferred technology, induced many rms to use foreign direct investment instead of developing local suppliers and transfer technology to them, something that would have required inscription in the National Registry of Technological Transfer (Martinez & Farber, 1994). On the other hand, the lack of an adequate trade agreement between the United States and Mexico implied that all inputs of Mexican origin generated duties upon entry in the United States. The same holds for those intermediate products that would have been modi ed to such extent as to be unrecognizable upon reentry in the United States. In consequence, and due to the importance of cost-reduction in maquila industry, the use of Mexican inputs was discouraged because of the import tari s. On the supply side, local rms have to pass often very strict and time-consuming processes (sometimes up to one and half years) of certi- cation and quality control to be accepted as a supplier of transnational corporations. In the meantime, the Mexican rm has to bear the cost of having the plant installed waiting for approval (Perez, 1992). On top of this, production of inputs required by maquiladora plants is technologically demanding and subject to continuous technological change. In the ECLAC-Mexico survey, comments were made on inputs of Mexican origin as being of substandard quality and high priced. Mexican producers were seen as unreliable with delivery deadlines, lacking scale for high-volume

11 MAQUILA, ECONOMIC REFORM AND CORPORATE STRATEGIES 1637 production, and not o ering the same nancial terms as their competitors abroad. Finally, the maquila program has not contributed signi cantly to the creation of local rms. In 1996, the maquila capital was predominantly controlled by US rms (64%), followed by Mexican and Japanese investors (21% and 5%, respectively). Regarding the Asian maquiladoras, free trade policy and particularly NAFTA have attracted those rms. During 1987±89 only 19 Japanese rms were under the maquiladora program. Later, during the NAFTA negotiation (1990±93), the number of Asian maquiladoras was expanded to 72: 38 from Japan, 12 from South Korea, 12 from Taiwan, and 3 from China (Mendiola, 1998). 4. CORPORATE STRATEGIES IN CENTRAL AMERICAN AND DOMINICAN MAQUILA Export Processing Free Zones and other provisions for maquila rms have existed in Central America and the Dominican Republic since the late 1960s. The San Bartolo Free Zone in El Salvador, the La Romana Free Zone in the Dominican Republic and the Temporary Admission Regime in Costa Rica were to some extent successful, resulting in the establishment of a few dozen rms. For di erent reasons, the activity did not take o in Guatemala, Honduras and Nicaragua (CEPAL, 1998a). Because political instability in the region hindered the expansion of assembly for export, the real take-o of assembly activities did not occur until The legal and institutional framework for maquila operations was modernized at the same time as the import substitution regime was dismantled and the NEM was adopted, roughly during 1983±90 (CEPAL, 1998a). An important caveat is in order here. In contrast to the situation in Mexico, in Central America and the Dominican Republic there does not exist a maquila program that covers all assembly activities. Firms may opt for di erent scal incentive regimes, each with speci c advantages and characteristics. Among the most common are the Export Processing Free Zones, the Temporary Imports Admission Schemes and the Draw-back mechanisms. 15 This situation hinders the gathering of uniform and comparable statistical information. With this in mind, data are presented in the following two tables on the present situation of the operations. Table 5 summarizes the participation of maquila exports in total exports. Although an exact percentage may not be available, it is a fact that most maquila activities involve apparel assembly. Noteworthy is the impressive growth of this indicator in Honduras, El Salvador, and to a lesser extent Guatemala. In Nicaragua, the assembly activity is still in its early phase, and in Costa Rica, the activity has stagnated in recent years. Relatively higher salaries and a new orientation of assembly industry in Costa Rica account for this trend. In 1996, maquila in Central America generated 235,000 direct jobs, which is 33.7% of manufacturing employment and 3.19% of total employment (see Table 6). In El Salvador and Costa Rica, the activity carries most weight in terms of percentage of manufacturing employment (59% and 45%, respectively). 16 In the Dominican Republic, maquila employment represents 38.5% of manufacturing employment and 6.5% of total employment. As a percentage of manufacturing value-added, the assembly activity is most important in Dominican Republic and Honduras. The origin of capital varies among Central American countries: in Costa Rica and the Dominican Republic it is mostly North American rms, in El Salvador and Guatemala rms are mostly of local capital, in Nicaragua it is East Asian capital and in Honduras it is evenly split. The ECLAC-Mexico survey reveals the principal productive characteristics of Central Table 5. Central America and Dominican Republic: maquila exports/total exports a Costa Rica n.a. El Salvador Guatemala Honduras Nicaragua n.a Dominican Republic n.a. n.a n.a. n.a. n.a. a Source: Department of Commerce and UNECLAC, Badecel.

12 1638 WORLD DEVELOPMENT American and Dominican maquila and the importance of technical change in rm strategies. Given the fact that the activity practically starts with the NEM, the analysis focuses on the present situation. Special attention is devoted to the rise of a new type of maquila activity in Costa Rica. The main competitive strategy of the maquila rms in the survey is to o er high quality, a result that con rms the nding that maquila is no longer just a matter of simple assembly against the lowest possible cost. Strong international competition in the apparel sector forces the rms to satisfy world-class standards. Regarding the participation of local personnel in strategic decision making, this is found to be negligible regarding the acquisition of inputs and raw material, the product mix and the sales and marketing operations. On the other hand, in the vast majority, local personnel are responsible for the organizational model, training and recruitment. Eighty- ve percent of maquila workers are reported to have primary and/or secondary education, 10% has technical training and 5% professional. Over 90% of rms in the survey have a very active policy of training of shop- oor workers, supervisors and managers. Courses are formal and are related to daily tasks and personal improvement. Furthermore, workers receive on-the-job-training. In the words of a manager: ``maquila is only possible if the human resource is adequately trained. Maquila implies, at international levels, productive and competitive workers.'' The tasks performed in Central American and Dominican maquila rms have evolved over time. The rst maquila rms (from the mid-1980s, particularly in Costa Rica) performed only sewing and packaging. To the degree that local plants acquired capacity to produce large volumes with e ciency and quality, other tasks were assigned to them, such as cutting and assembly of accessories. It should be born in mind that a factor that has in uenced this conformation of so-called total package production is the worldwide tendency to concentrate production phases in large plants that perform a variety of manufacturing processes (CEPAL, 1998a). Maquila rms in the survey do not have direct participation in product design. They receive speci cations and designs, and only occasionally will they perform tests. The production process is characterized by long lines of around 50 workers, and less frequently, modules and work groups. Paramount is simple tasks on traditional sewing machines, although in a few occasions automated processes may be observed with sophisticated machinery for numerical control of production processes. Within the plant there are basic capabilities of machine maintenance and repair, but if major problems arise, the machine provider or corporate headquarters has to be contacted. It is in the organization of the rm and the production process that the most evident tendency of modernization was encountered. The programs employed are designed to form an organizational culture based on quality and e ciency. Noteworthy are above all the implementation of work groups, programs to increase industrial safety, just-in-time and total quality control, as well as the elaboration of manuals, procedures and written internal norms. Following the classi cation applied to distinct degrees of technological evolution of Mexican maquila, in general maquila in Central America and the Dominican Republic may be classi ed as pertaining to the second generation. This is based on three observations: the importance of quality, the characteristics of the organization of the plant and the production process, and the presence of automated machinery. In the particular case of Costa Rica, from the mid-1990s on a series of new maquila rms have set up shop that have a di erent technological complexity from traditional apparel assembly. The relatively high salaries compared to the rest of Central America and Mexico 17 and the new strategy followed by the Government of Costa Rica noticeably in uenced apparel maquila. Acting upon the belief that apparel maquila is not what the country needed, in the administration of President Figueres (1994±98) an aggressive policy was developed to attract foreign direct investment for the establishment of assembly operations in the microelectronics sector. The e ort had as its most important result the decision by INTEL to build a microchip assembly and inspection plant near San Jose. In May 1998, the ECLAC-Mexico survey team visited ve assembly plants of microelectronic products around San Jose, all of US origin. It was found that the organizational model, training, and recruitment of personnel are the responsibility of local managers. In contrast, decisions about acquisitions of inputs and product mix are the responsibility of corporate headquarters.