The Role of Maquiladoras in Mexico's Export Boom -- Gordon H. Hanson
The Role of Maquiladoras in MexicoÂ¡Â¯s Export Boom
Gordon H. Hanson
University of California, San Diego
and National Bureau of Economic Research
Prepared for the Conference, Â¡Â°Prospects for Industrial Parks in the Palestinian Territories,Â¡Â± Rice University, July 26-27, 2002.
Over the last two decades, Mexico has dramatically opened its economy to foreign trade and investment. The country has converted itself from an inward-oriented economy to one in which export production is the main source of economic growth. In 2000, the share of international trade in MexicoÂ¡Â¯s GDP was 32%, up from 11% in 1980. The most dynamic exporters in Mexico are in-bond assembly plants, known as maquiladoras. These plants import parts and components from abroad, assemble the inputs into final goods, and then export their output. They are most active in the electronics, auto parts, and apparel industries. Maquiladoras main point of contact with the Mexican economy is through hiring labor. They purchase few inputs in Mexico and sell virtually none of their output domestically. The United States is the primary source for their inputs and the primary destination market for their sales. While all maquiladoras contract with foreign firms to obtain inputs and distribute outputs, not all are subsidiaries of multinational enterprises. Many are Mexican-owned facilities that deal with multinationals through arms-length transactions.
Between 1990 and 2002, real value added by the maquiladora industry grew at an astounding annual average rate of 10%. To put this growth in perceptive, over the same period real GDP in Mexico expanded by an annual average rate of only 3%. Since 1984, employment in maquiladoras has risen from 180,000 workers to 1.1 million workers, or to over one-quarter of MexicoÂ¡Â¯s total manufacturing labor force. By 2000, the maquila sector generated 48% of MexicoÂ¡Â¯s exports and 35% of the countryÂ¡Â¯s imports. These plants remain concentrated in Mexican states along the Mexico-U.S. border, which in 2002 accounted for over 80% of total maquiladora employment.
In this paper, I address two important questions about MexicoÂ¡Â¯s maquiladoras and their role in the countryÂ¡Â¯s export boom. First, what accounts for their success? Four factors help explain why maquiladoras dominate export production in Mexico: (i) the ability of multinational firms to fragment production across borders through global outsourcing, (ii) MexicoÂ¡Â¯s low wages, (iii) trade policies in Mexico and the United States initially that gave maquiladoras special advantages in exporting to the U.S. market, and (iv) MexicoÂ¡Â¯s proximity to a large, high-wage economy in the United States. Low policy barriers to trade, low wages by regional or international standards, and access to large markets appear to be necessary conditions for maquiladoras to expand.
Second, are there disadvantages to relying on maquiladoras to spur economic growth? I identify three potential disadvantages to a maquiladora-led model of development: (i) maquiladoras tend to specialize in a small number of volatile manufacturing industries, (ii) maquiladoras tend to be very sensitive to business cycles in other countries, and (iii) maquiladoras are footloose establishments that can easily relocate to another country if local costs rise. These disadvantages suggest that while maquiladoras can generate rapid export growth, they may also increase the sensitivity of an economy to global shocks. They may expand rapidly when times are goods, but they may also contract sharply when times are bad.
Over the last two decades, Mexico has dramatically opened its economy to foreign trade and investment. Mexico began this process of liberalization in the 1980Â¡Â¯s by unilaterally lowering barriers to imports and removing restrictions on multinational firms. It consolidated these reforms in the 1990Â¡Â¯s by signing the North American Free Trade Agreement (NAFTA) with Canada and the United States. The country has converted itself from an inward-oriented economy to one in which export production is the main source of economic growth. By 2000, the share of international trade in MexicoÂ¡Â¯s GDP was 32%, up from 11% in 1980. Foreign direct investment (FDI) has helped fuel MexicoÂ¡Â¯s export growth, quadrupling from less than 1% of GDP in the early 1980Â¡Â¯s to over 3% of GDP in the late 1990Â¡Â¯s (Hanson, 2002).
The most dynamic exporters in Mexico are in-bond assembly plants, known as maquiladoras. These plants import parts and components from abroad, assemble the inputs into final goods, and then export their output. They are most active in the electronics, auto parts, and apparel industries. Maquiladoras main point of contact with the Mexican economy is through hiring labor. They purchase few inputs in Mexico, beyond packing materials and the water and power needed to keep factories running, and sell virtually none of their output domestically. The United States is the primary source for their inputs and the primary destination market for their sales. While all maquiladoras contract with foreign firms to obtain inputs and distribute outputs, not all are subsidiaries of multinational enterprises. Many are Mexican-owned facilities that deal with multinationals through arms-length transactions.
MexicoÂ¡Â¯s recent export boom would not have been possible without maquiladoras. Between 1990 and 2002, real value added by the maquila sector grew at an astounding annual average rate of 10%. During this same period, employment in maquiladoras rose from 460,000 workers to 1.1 million workers, or to over one-quarter of MexicoÂ¡Â¯s total manufacturing labor force. By 2000, the maquila sector generated 48% of MexicoÂ¡Â¯s exports and 35% of the countryÂ¡Â¯s imports. These plants remain concentrated in Mexican states along the Mexico-U.S. border, which in 2002 accounted 83% of maquiladora employment (down from 95% in 1990).
In this paper, I address two important questions about MexicoÂ¡Â¯s maquiladoras and their role in the countryÂ¡Â¯s export boom. First, what accounts for their success? While MexicoÂ¡Â¯s trade and investment reforms in the 1980Â¡Â¯s and 1990Â¡Â¯s raised incentives for export production, there are many other types of exporters that conceivably could have answered the call. I argue that four factors account for why maquiladoras have dominated export production in Mexico: (i) the ability of multinational firms to fragment production across borders through global outsourcing, (ii) MexicoÂ¡Â¯s low wages relative to the rest of North America, (iii) trade policies in Mexico and the United States initially that gave maquiladoras special advantages in exporting to the U.S. market, and (iv) MexicoÂ¡Â¯s proximity to a large, rich economy in the United States. Low policy barriers to trade, low wages by regional or international standards, and access to large markets appear to be necessary conditions for maquiladoras to expand.
The second question is, are there disadvantages to relying on maquiladoras to spur economic development? Most developing countries that have achieved high rates of economic growth in recent decades, including China and the nations of East Asia, have had dynamic export sectors. In some of these countries, export production has been dominated, at least for a time, by maquiladora-type establishments. But in others, export production has been more diversified. I identify three potential disadvantages to having a large maquila sector, all of which relate to the volatility of national income: (i) maquiladoras tend to specialize in a small number of volatile manufacturing sectors, (ii) maquiladoras tend to depend on firms in a small number of foreign markets for sales, and (iii) maquiladoras are footloose establishments that can easily relocate to another country if local costs rise. These disadvantages suggest that while maquiladoras can generate rapid export growth, they may also increase the sensitivity of an economy to global shocks. Maquiladoras may be more responsive than other exporters to changes in their costs or in the demand for their output. They may expand rapidly when times are goods, but they may also contract sharply when times are bad.
To provide context for the discussion of maquiladoras, in this section I give a brief overview of MexicoÂ¡Â¯s policy reforms and the structure of MexicoÂ¡Â¯s labor force.
Policy Reform in Mexico. The last several decades have not been a quiet period in the Mexican economy. Since 1980, the country has had three currency crises, periodic bouts of high inflation, and several severe macroeconomic contractions. The reform of the countryÂ¡Â¯s trade and investment policies has been, in part, a response to this economic turmoil. Following a balance of payments crisis in 1982, the country eased restrictions on maquiladoras, which lead to a large increase in export processing trade with the United States. In 1985, Mexico broadened its opening to trade by joining the General Agreement on Trade and Tariffs (GATT), which entailed cutting tariffs across the board and eliminating many non-tariff barriers. In 1984, import licenses covered 92% of national production, average tariffs were 24%, and there were export controls on 85% of non-petroleum exports. By the end of 1987, the Mexican government had abolished export controls, reduced import-license coverage to 25% of production, and cut average tariffs to 12% (Hanson, 1998). In 1989, Mexico extended liberalization to foreign investment by easing restrictions on the rights of foreigners to own assets in the country. NAFTA consolidated and extended these reforms and tied them to reciprocal access to the U.S. and Canadian markets. Mexico is now as closely tied to the North American economy as it has been at any point in its history. In 2000, the country sent 89% of its exports to and bought 73% of its imports from the United States (Hanson, 2002).
Population and Employment in Mexico. Table 1 gives summary statistics on working age adults in Mexico in 1990 and 2000. Mexico is a young country. For adults 16-65, the average age is 33 years. The vast majority of adults are literate and education levels, while low by developed country standards, have been rising over time. The fraction of adult men who had completed 12 or more years of education rose from 19% in 1990 to 25% in 2000. Figures for women are similar. In the United States, by way of contrast, over 85% of adults have completed 12 or more years of education.
MexicoÂ¡Â¯s population is concentrated in the center of the country. In 2000, 57% of the adult population lived in the capital region, in which Mexico City is located, or in surrounding central states. During the 1990Â¡Â¯s, the share of the population in the border region, in which most maquiladoras are located, rose slightly from 17% to 18%. This small increase in the border population is remarkable, given the dramatic growth in maquiladora employment and in the population of large cities, such as Tijuana and Ciudad Juarez, that occurred in the region. Border employment has grown primarily through firms attracting workers from rural areas or neighboring states within the border region. Migration flows between regions of Mexico have been small.
Mexico is an increasingly urban country. The share of the population living in cities with more than 100,000 inhabitants rose from 47% in 1990 to 51% in 2000. This increase occurred entirely at the expense of rural areas (localities with less than 2,500 inhabitants). Changes in MexicoÂ¡Â¯s land tenure system during the 1990Â¡Â¯s have helped increase the outflow of population from the countryside.
Wages in Mexico are low by North American standards. In 2000, average hourly wages were $1.80 for men (in current dollars). Real wages in Mexico (deflated by MexicoÂ¡Â¯s consumer price index) fell in the 1990Â¡Â¯s, due in large part to a severe recession in 1995 (in which real GDP contracted by over 6% in a single year). To put these wages in context, it is helpful to compare them to wages for U.S. workers with similar skill levels. Recent Mexican immigrants in the United States (individuals with less than ten years in the country) are an obvious comparison group. In 2000, average hourly wages for recent male Mexican immigrants in the United States were about $8.50. To a first order of approximation, wages in the United States are 4-5 times wages in Mexico.
Another important feature of Mexican wages is their variation across regions. Mexican states that border the United States have the highest wages in the country. In 2000, (controlling for observable characteristics of workers) average hourly wages for men in the border region were 20% higher than wages in the next tier of northern states, 27% higher than wages in central states (including Mexico City), and 58% higher than wages in southern states (where Chiapas is located). What is more, the border wage premium increased during the 1990Â¡Â¯s. These large regional wage differentials make the absence of regional migration flows in Mexico all the more surprising.
Table 2 describes the distribution of employment across sectors in Mexico. During the 1990Â¡Â¯s, manufacturing held stable at around 20% of the countryÂ¡Â¯s labor force. Table 3 describes the distribution of industry employment across regions. During the 1990Â¡Â¯s, the share of national manufacturing employment located in the border region increased from 22% to 26% (for men). Manufacturing jobs are clearly mobile across regions in Mexico, even if the workers themselves are not.
3. Explaining MexicoÂ¡Â¯s Maquiladora Boom
Since the early 1980Â¡Â¯s maquiladoras in Mexico have been the most dynamic sector in the Mexican economy. Figure 1 shows employment and real value added in maquiladoras by month from 1984 to 2002. Employment grew from 180,000 workers in 1984 to 1.3 million workers in 2000 before declining to 1.1 million workers in 2002. Real value added (in 1994 dollars) grew from $330 million in 1990 to $1.5 billion in 2000 before declining to $1.2 billion in 2002. To put this growth in perceptive, while real value added in maquiladoras expanded by 10.4% per year between 1990 and 2002 real GDP in Mexico expanded by only 2.9% per year over the same time period. What explains MexicoÂ¡Â¯s maquiladora boom? In this section, we identify several key factors that help account for the industryÂ¡Â¯s rapid growth.
3.1 Global Outsourcing and Mexican Wages
Maquiladoras are the result of multinational firms outsourcing assembly operations to Mexico. For outsourcing to be an efficient strategy, four conditions must hold: (i) it must technologically feasible for firms to locate different stages of production in different countries, (ii) the intensity with which labor, capital, and other factors are used in production must vary across production stages, (iii) wages and other factor costs must vary across countries, and (iv) transport and other trade costs must not be too high.
To understand the logic behind outsourcing, consider a stylized model of the production of television sets, a major output of MexicoÂ¡Â¯s maquiladora industry. LetÂ¡Â¯s suppose that we can break down TV production into four distinct stages, design, parts production, assembly, and marketing:
Design Â¡ú Parts Production Â¡ú Assembly Â¡ú Marketing
(skills) (skills and capital) (labor) (skills)
This organization of production is common to many semi-durable consumer goods industries, including other consumer electronics items, auto parts, apparel, and footwear (Feenstra and Hanson, 2002). Each of these stages differs considerably in its factor intensity. Design requires engineers who draw up blue prints and construct prototypes, making this stage relatively intensive in the use of skilled labor. Parts production involves constructing picture tubes, electronic circuitry, and sound systems, which require moderately skilled labor, sophisticated machinery, and clean manufacturing facilities. This makes this stage relatively intensive in the use of semi-skilled labor and physical capital. Assembly is the most labor-intensive stage, requiring workers with no more than a basic education. Marketing is also likely to be relatively skill-intensive, especially if firms require sophisticated advertising campaigns to sell their TVs.
The logic behind global outsourcing is to locate each production stage in the country where it can be performed at least cost (Grunwald and Flamm, 1985). Firms will locate design and marketing in countries with abundant supplies of skilled labor, such as the United States, Europe, or East Asia. Parts production may also go to these countries or to countries one step down in terms of their supply of skilled labor. Assembly, in which unskilled labor is the primary input, will go to low-wage countries.
Low wages in Mexico are key to its ability to attract multinational firms wishing to set up export assembly operations. Figure 2 shows average hourly wages (in 2002 US dollars) for production workers in Mexican maquiladoras in border states and in other states for the period 1991-2002. In Mexico, the maquiladora boom occurred during a period of flat wage growth. During the 1990Â¡Â¯s, when employment in maquiladoras more than doubled, wages fluctuated between $1.50 and $1.75 an hour. A large increase in wages, as I discuss in section 4, would be likely produce an exodus of maquiladoras from the country. Indeed, after wages did increase in the early 2000Â¡Â¯s, maquiladora employment fell sharply (though other forces were also at work).
Outsourcing assembly to Mexico or another low-wage country depends on firms being able to break up the production process and locate different stages in different countries (Feenstra and Hanson, 1997). This is feasible in industries in which production stages are technologically distinct and in which inputs and outputs can be transported between countries at relatively low cost. Electronics, auto parts, apparel, footwear, and other semi-durable consumer goods fit this profile. But many other industries, such as petrochemicals, cement, or commercial aircraft do not. This is one reason that maquiladoras are highly concentrated in specific industries. Table 4 shows the share of national maquiladora employment by industry in Mexico for 1980-2000. In 2000, just three industries, electronic accessories, auto parts, and apparel, accounted for 66% of employment in the sector. Over the last two decades, as the industry has grown considerably, there has been no apparent tendency towards greater diversification.
Beyond the technological feasibility of separating production into separate stages, outsourcing requires firms to ship inputs and outputs back and forth across national borders with great frequency. One issue for firms is the ease with which they can communicate with plant managers in other countries. Recent advances in information and communication technology surely makes interacting with remote manufacturing operations less costly than in the past. But of greater importance are the more mundane costs associated with policy barriers to trade and with shipping goods between countries. Trade and transport costs have been very important in shaping MexicoÂ¡Â¯s maquiladora industry. I discuss these factors in the next two sections.
3.2 Mexican and U.S. Trade Policies
Mexico created its maquiladora sector at a time when the country was largely closed to international trade. The sector thus began as an export enclave within an inward-oriented economy. After Mexico opened its economy to trade, maquiladoras lead the ensuing export boom. This was due in part to U.S. trade policies that gave maquiladoras privileged access to the U.S. market.
In the 1940Â¡Â¯s, Mexico adopted a strategy of import substitution industrialization. The government decided to protect the economy from imports in order to encourage the development of domestic industry. To import most manufacturing products, firms had to obtain a license from the government and had to pay moderate to high import tariffs. The imposition of import licenses and tariffs had two important effects: it gave firms a captive domestic market and it raised the costs to firms of obtaining the inputs they would need to produce many types of exports. Both policies had the effect of reducing the incentive to export. On top of this, in many industries, the government limited exports to a certain percentage of national production.
In 1965, Mexico softened its import substitution strategy by allowing the creation of maquiladoras as part of its Border Industrialization Program (Hansen, 1981). The program permitted firms to import free of duty the inputs, machinery, and parts they would need for export assembly operations. Maquiladoras, as they plants came to be known, were required to export all output. To ensure firms abided by this rule, they were required to buy a bond equal to the value of their imports that would be returned to them once they had exported all their imported inputs in the form of final goods (hence the term in-bond assembly plants). In contrast to other firms in the country, maquiladoras could be 100% foreign owned. Initially, maquiladoras were required to locate within 20 miles of an international border or coastline. In 1972, the government relaxed these rules and allowed maquiladoras to locate throughout the country.
Initially, there was little activity in the maquila sector. In 1975, after policies to create the sector had been in effect for ten years, total employment in maquiladoras was only 62,000 workers. One constraint on their growth was the maquiladora program itself. Many legal rules governing the sector were poorly defined and there was the perception that Mexican Customs complicated the process of importing inputs, obtaining foreign currency, and repatriating profits. All of this changed in 1982, with the onset of a severe balance of payments crisis in Mexico. The government issued a series of decrees that streamlined the bureaucracy governing maquiladoras, making them easier to establish, own, and operate. It was in 1983 that the sector began its rapid expansion.
Prior to 1982, there were other factors limiting the growth of maquiladoras, as well. While maquiladoras could import inputs duty free, existing trade barriers still gave Mexican firms a captive domestic market. These barriers allowed domestic producers to absorb resources that under free trade would have gone into export production. In the early 1980Â¡Â¯s, policy incentives against exports began to disappear. A devaluation of the peso in 1982 helped make Mexican exports more competitive internationally and MexicoÂ¡Â¯s trade reform in 1985 helped remove the artificial incentive to produce for the domestic market. These policies further helped the maquiladora industry expand.
After Mexico removed policy barriers that had kept export production artificially low, why were maquiladoras, and not other firms, the ones to lead the export boom? One reason is the U.S. offshore assembly program (OAP) (Hanson, 2000). The U.S. OAP permits the duty-free return of domestically-manufactured components that have been processed in another country. The importing firm is required to pay import duties only on the value added abroad. OAPs reduce the cost to U.S. firms of moving assembly operations abroad. A U.S. firm that makes components, ships them to a plant in Mexico for assembly, and then reimports the finished good will, between the two countries, only pay import duties in the United States on the value of Mexican labor and raw materials used in assembly. In contrast, if the firm were to buy inputs in Mexico and have them assembled there, it would pay U.S. import duties on the entire value of the good. Mexico is the third largest supplier of OAP imports, accounting for 17% of total U.S. OAP imports in 1990.
U.S. trade policies initially gave maquiladoras an advantage over integrated Mexican producers Â¨C firms that produced their own inputs and assembled these inputs into final outputs Â¨C in exporting to the U.S. market. NAFTA ended this special status for maquiladoras by giving all Mexican firms duty free access to the U.S. market. Many predicted that NAFTA would bring an end to MexicoÂ¡Â¯s maquila sector. How would these firms survive if they no longer benefited from special access to the United States? As Figure 1 suggests, the implementation of NAFTA in 1994 did little to stunt the growth of the maquila sector. In a purely legalistic sense, NAFTA did mean the end of the maquiladora regime: it eliminated the "in-bond" arrangement under which maquiladoras operated. Given Mexico's low wages, however, it is likely that the country will continue to specialize in the assembly of manufactured goods for North America.
The more pressing question is which country will produce the components that maquiladoras assemble. The pre-NAFTA pattern of trade between the United States and Mexico tells us something about each country's comparative advantage (Hanson, 2000). Prior to NAFTA many goods, including TVs, motor vehicle parts, and apparel, that were produced from U.S. components and assembled in Mexico were consumed in both the United States and Mexico. Even with the pre-NAFTA tariff disadvantage in the Mexican market, U.S.-made components were cheaper than Mexican-made components. The abolition of trade barriers should have only strengthened the comparative advantage of the United States in components production. The most likely scenario, then, is that NAFTA will cause Mexican assembly plants and U.S. components producers (and not Mexican components producers) to expand. Barring competition from other low-wage countries, NAFTA is likely to have strengthened MexicoÂ¡Â¯s maquiladora industry.
3.3 Proximity to the North American Market
In its original conceptualization, supporters of the maquiladora program envisioned a "twin plant" arrangement, in which a facility located in a U.S. border city would manufacturer components and a facility located in the neighboring Mexican border city would assemble the components into a finished good (Sklair, 1989; Wilson, 1992). A common management team in the United States would run both plants. The physical proximity of the plants would minimize transport costs in shipping inputs to Mexico and outputs to the United States. While few firms fully implemented the twin-plant scheme, proximity to the U.S. border has played an important role in the industry.
What is striking about the location of maquiladora operations in Mexico is the extent to which they continue to concentrate on the border. Figure 3 shows the share of national employment in maquiladoras located in Mexican states on the Mexico-U.S. border over the period 1992-2002. While this share has declined over time, it is still over 80%. This is surprising given that wages in border states are much higher than in the rest of the country, as shown in Figure 2, and that policy restrictions confining maquiladoras to border states were lifted thirty years ago. Maquiladoras are clearly willing to pay a premium to locate close to the United States. This might not be surprising if workers in border states were more productive than workers elsewhere in the country. But available evidence suggests this is not the case. Figure 4 shows real value added per worker (in 1994 dollars) for maquiladoras in border states and maquiladoras located elsewhere in the country. Plants in border states do not appear to have higher labor productivity than plants in other states.
There are at least three important advantages for maquiladoras of being close to the United States. First, obviously, is transport costs. The closer are plants to the U.S. market the lower are costs of shipping parts and components to Mexico and of shipping finished goods back to the United States. Second, it allows managers based in the United States to monitor production operations in Mexico. U.S. and Asian multinationals appear to have found it difficult to persuade managers to live in Mexico. By locating maquiladoras in Mexican cities on the Mexico-U.S. border, managers can oversee their operations but still live in U.S. cities, such as San Diego, El Paso, or McAllen. Third, there appear to be benefits to agglomerating maquiladora operations in specific locations. The larger the number of maquiladoras in a city, the larger the number of workers who come to the city seeking work, helping plants deal with chronic high rates of labor turnover in the industry. Agglomeration also raises incentives for developers to provide services, such as creating industrial parks and linking plants to municipal water, power, and transportation services. Once plants became established in the border region, benefits to agglomeration may have helped keep them there despite other disadvantages.
It is worth noting that as export manufacturing has expanded in Mexican border cities, a wide range of activities have expanded in U.S. border cities (Hanson, 1996 and 2001). Smaller U.S. border cities function as intermediaries in U.S.-Mexico trade, providing transportation and distribution services. It is not too surprising that these border intermediary services have expanded with Mexican exports to the United States. Some amount of transhipment inevitably occurs at border crossings, which creates local demand for related activities. Larger U.S. border cities are the base for manufacturing operations of U.S. firms that locate in the border region. Manufacturers in El Paso, McAllen, and Brownsville supply maquiladoras in Mexico with parts and components for the assembly of apparel, automotive products, and electronic appliances. Much of this production utilizes generic techniques, such as metal stamping and plastic injection molding, for a wide variety of components hat are used in autos and electronics. These manufacturing activities are closely tied to the product assembly performed by maquiladoras.
4. Disadvantages of Maquiladora-Led Development
Many view MexicoÂ¡Â¯s maquiladora industry as a huge success. The industry has lead a prolonged export boom in Mexico, helping integrate the country into the world economy. Wages in the industry are higher than wages in other manufacturing activities, spreading the benefits of globalization to production workers with relatively modest skills. The industry also offers more employment opportunities for women than other manufacturing activities, making it the vehicle through which many women have entered the manufacturing labor force. The main complaints about the industry have been that Mexico did not plan well enough for its growth (Wilson, 1992). After two decades of nearly continuous expansion, border cities have become extremely congested and lack adequate housing, infrastructure, and basic services.
Recently, other defects of the maquiladora industry have come to light. Until 2000, MexicoÂ¡Â¯s experience with the industry, which effectively began in the early 1980Â¡Â¯s with the governmentÂ¡Â¯s streamlining of regulations governing maquiladoras, had coincided with a prolonged economic expansion in the United States. Mexico simply hadnÂ¡Â¯t lived through a period in which the U.S. industries that are the main source of demand for maquiladora output suffered