This Research Bulletin has been published in S Sassen (ed) (2002) Global Networks, Linked Cities New York, London: Routledge, 145-82.
With a population of 16.7 million people in 1995, Mexico City2 is one of the largest urban agglomerations in the world. This leads into thinking of it merely as a "Third World" mega-city. However, this is a simplified picture, because beyond the large size new dynamics of urbanization may be at work, which are related to economic globalization. Though these new dynamics are far more easily seen in "First World" metropolis - among other things because we do not have the mega-city syndrome - they are also present in "Third World" cities. Unfortunately, in the case of "Third World" cities the new elements in urban processes tend to be submerged under the size of the population, which impedes to grasp the growing convergence between major cities in the centers and in the peripheries of the world economy. To overcome this deficit I focus on the global integration of Mexico City. My main argument is that it is not quantity - in terms of population size - which shaped Mexico City's destiny in the last two decades, but the changing quality of relations with the "national economy" on the one hand, and with the "world economy" on the other hand. I maintain that the city's development was strongly biased by the radical shift in the country's economic strategy towards foreign markets, because this shift implied a revalorization and reordering of economic activity, of space, and of capital-labor relations. Thus, the emphasize of this article is on a) transformations of Mexico City which occurred with the deeper integration of Mexico into the international division of labor; and b) on the specific role of Mexico City in this process of globalization. By concentrating on global embeddedness, this article aims to contribute to our knowledge of Mexico City, which derives from rich studies on the city's economic, spatial and social development (see, for example, Garza 1985; Ward 1998).
My approach is based on the global or world cities research, which re-emerged in the 1980s and which has since then created a theoretically and - to a lesser extent - empirically compelling body of literature (see, for example, Friedmann 1986; Sassen 1991; Knox/Taylor 1995; Lo/Yeung 1998). In short, the argument is that globalization processes give rise to a new form of centrality, in which a certain number of cities emerge as key places. These global cities articulate larger regional, national and international economies into the global economy, and by doing so they serve as nodal points where the flows of money, information, commodities and migrants intersect and from where they are directed. Concentrating command and control functions, global cities are both production sites and trade places for specific "goods", namely financial and other advanced producer services, which are essential for global integration. Finally, global cities are linked to each other by flows of money, information, commodities and migrants, creating thus a cross-border network of cities. The emergence of a global urban system alters the geography of the world system (perceived traditionally as a collection or hierarchy of nation-states), because it operates both through nation-states and by bypassing their boundaries.
Until recently, global city research gave priority to metropolis in the centers of the world system. Most - not all - studies on "Third World" urbanization used to place cities in a nation-state framework3. In addition, they showed a strong bias towards demographic questions, the "urban primacy" and urban problems. In recent years, however, the call to put the conceptual and theoretical relationship of core-world cities to the mega-cities of the peripheries on the top of the research agenda become louder (Friedmann 1995, 43; Knox 1995, 16; Sassen 1996, 14-16; Gugler 1999, 1f; Taylor 1999b, 2). There are, in fact, good reasons to expand global city research to "Third World" cities. First, there are serious limitations in the conceptual framework and database for adequate assessments of mega-cities. The deficiencies can be illustrated by the simple fact that the main cities in Latin America, Africa and Asia are still called simply "mega-cities", which is of course a quantitative definition, compared to the qualitative approach of "world" or "global city". Second, a country's position in the international division of labor influences its economic and social development. Accordingly, urbanization in the "Third World" is linked to broader dynamics such as colonialism, merchant capitalism, dependent industrialization or the crisis of the world economy (Chase-Dunn 1985; Gilbert 1992; Feldbauer et al. 1993). Third, if the postulate is taken seriously that today's world economy is shaped as a network, and if cities are, as argued, nodal points in this network, then the major cities of the "Third World" should be expected to be key places for globalization, too. From this perspective, a city like Mexico City presumably is the place where the integration of economic activities, territories and societies of a "Third World" country such as Mexico into the global system is organized, controlled and managed. Thus, one can conclude that the mega-cities of the periphery are an essential element of the cross-border urban system (Parnreiter 1998b).
In the following sections I will analyze major developments in Mexico City in light of global embeddedness. Though results are still tentative, they show that the deep crisis of the 1980s, the recovery and restructuring of the urban economy in the 1990s, the turn towards an area of net-emigration, and impoverishment and social polarization are closely linked to the particular way in which Mexico and its capital have been integrated into processes of globalization. However, it is not only that forces operating on a global scale have affected Mexico City. Certain parts of the city assume more economic activities that are of global - or at least of regional - reach. Thus, due to globalization, and as part of it, Mexico City turns from a national metropolis into a "hinge" between the Mexican and the global economy. By doing so, Mexico City contributes to the "production" of globalization.
MEXICO IN THE GLOBAL ECONOMY
Integration of Mexico into the world system is by no means a recent phenomenon. Since the Spanish conquest over externally stimulated and pushed modernization during the rule of Porfirio Díaz (1876-1911) to massive indebtedness at the end of import substitution Mexican destiny has been shaped to a good deal by factors beyond its territory. However, in the two decades following the debt crisis in 1982, Mexico's integration into the international division of labor became ever deeper and it changed its form. By joining GATT (today: WTO) in 1986 and the North American Free Trade Area (NAFTA) in 1994, the country' government institutionalized its course towards globalizing Mexico.
The collapse of import substitution, marked by the debt crisis, forced Mexico to a radical change of its economic strategy. After the bankruptcy in 1982, the Mexican government had to find ways to guarantee the payment of the interest (and, to a smaller extent, the re-payment of the debt) and to restore public finance. That meant, in the context of a changing international environment, the deregulation and opening of the national economy, its re-orientation towards external markets, and the dismantling of labor standards and social policies. In order to create opportunities for highly profitable investment for both foreign capital owners and national business élites, Mexican governments implemented neoliberal policies. Beginning with the presidency of Miguel de la Madrid (1982-1988) and continued under Carlos Salinas (1988-1994) and Ernesto Zedillo (1994-2000), a good part of the country's economy and labor market were deregulated, the financial sector liberalized, land reform halted, tariff protections dismantled, state-owned enterprises privatized, manufacturing and agriculture re-orientated towards foreign markets, wages and social transfers reduced, and markets opened up to foreign products, firms and investors. All these changes show that globalization did not just happen to Mexico. Rather, it was produced. Though certainly under pressure of "global players", the Mexican state - or the élite controlling it - was a crucial agent in designing neoliberal policies. Globalization thus implies the transformation of state activities in order to assert the interests of a specific set of private sector firms, namely those operating on a global scale. In other words: Economic globalization is a intentionally chosen strategy to restructure economic activities socially, spatial and sectoral in order to restore and expand profits.
Due to the new development strategy, some economic indicators show favorable results. Annual exports grew more than fivefold between 1982 and 1999, driven mainly by export manufacturing. A growing part of all Mexican exports originates in the maquiladora industry4 (1980-1989: 21%; 1990-1998: 42%), which expanded its exports twelvefold between 1982 and 1998. Annual production of these assembly plants grew eightfold since 1980, and so did employment, which amounted to 1.2 million workers at the end of 1999. Another result of the greater opening of Mexico towards the world market is that foreign capital flew in unprecedented dimensions to the country. Between 1989 and 1998, more than 80 billion dollars were invested in Mexico as Foreign Direct Investment (FDI), and the same amount flew in as portfolio investment (1989-1996). Mexico was - behind China - second among the "emerging markets" in receiving FDI in the first half of the 1990s, but is surpassed by Brazil since 1996. The stock of accumulated FDI grew more than eightfold between 1980 and 1997. Inflation, which amounted to an annual average of 75% in the 1980s, sank to less than 20% (annually) in the 1990s. The financial deficit decreased notably, thanks to privatization, fiscal reform, reduction of interest rates according to the "Brady Plan", and a deep cut in public expenditures (Bancomext 1999; La Jornada 17.1.2000; de Mateo 1998, 19f; Bendesky 2000; Dussel Peters 1999, cuadro 1, 2; Red Mexicana 1997, 42; Chudnovsky et al. 1999, 17-21; CEPAL 2000, 43; Bancomext 2000; Dussel Peters 1997, 150).
Despite some favorable results, globalization and neoliberal transformation did not benefit the Mexican economy as a whole. GDP grew by only 2% on an annual average (1982-1999), which is much lower than during the era of import substitution. GDP per capita did not grow at all. The balance of trade, which was positive in the 1980s, turned to be strongly negative in the 1990s, and the current account, which was already negative in the 1980s, deteriorated markedly in the 1990s. Both public and private foreign debt increased notably, amounting together to 172 billion US dollars in 1996 (own calculations, based on INEGI 2000a; Dussel Peters 1997, 150; Bancomext 1999; INEGI 2000b).
Thus, globalization did not meet the economic expectations raised by the government. Even worse, it could not alleviate the social crisis. Contrary to the rhetoric of neoliberal transformation, which promise growth and wealth resulting from globalization, living conditions for most Mexicans deteriorated in the last two decades. According to the World Bank, more than 40% of the population live in poverty, whereas Mexican researchers assess that no less that 45% of the country's population live in extreme poverty and further 25% in poverty. Impoverishment is, on the one hand, due to the decay of wages - in 1999, real minimum wages amounted to less than a third of its value in 1980. On the other hand, the deterioration of the standard of living is linked to the spread of precarious working conditions - between 50% and 60% of the economically active population work in the informal economy. As a consequence of wage decay and erosion of working standards, the average per capita income of Mexican households decreased by 10% between 1984 and 1996. However, not the whole population lost. While 90% of all households had less income in 1996 than in 1984, the richest tenth increased its per capita income by more than 7%. The growing polarization is shown by the fact that the per capita income of the richest tenth was 3 times bigger than the average income and nearly 20 times bigger the income of the poorest tenth in 1996, up from 2.5 and 14, respectively, in 1984 (Delgado Selley 2000; Dussel Peters 1997, 150; El Universal, 31.3.1997; El Financiero 21.1.2000; OECD 1999; CONAPO 1999, 153).
ECONOMIC, SOCIAL AND DEMOGRAPHIC TRANSFORMATIONS IN MEXICO CITY
In order to relate transformations in Mexico City ("Zona Metropolitana de la Ciudad de México"; ZMCM) to globalization, I chose four issues: The deep crisis of the early 1980s, the following restructuring and recovery, the impoverishment of large segments of the population and a deepened social polarization, and the transformation of the ZMCM into an area of net-emigration. Before discussing them, I shall briefly sketch out some aspects of recent developments.
Until the 1970s, Mexico City was the unchallenged epicenter of the national economy. However, the collapse of import substitution hit the ZMCM harder than other Mexican cities. Its share in the national GDP decreased from 36% to 32% in only five years (1980-1985), due mainly to the breakdown of the manufacturing sector. Industrial output shrank in absolute terms (-5.8% annually between 1980 and 1985), and its share in Mexican manufacturing fell from 48.6 to 32.1%. It is important to note that virtually all of the decline was caused by the decay of the Federal District's manufacturing sector. While various of the big companies shifted their production facilities either to nearby states in the central region or to the Northern part of Mexico, many of the medium and small local firms, which were not able to compete under conditions of free trade, had to close down (Pradilla Cobos 1997, cuadro 2; Garza/Rivera 1994, 13f; INEGI, various years [a]). Additionally, the Federal District lost importance as a national center of economic decision-making. While in 1982 287 of the Mexican "Top 500" companies were headquartered there, this number declined to 145 by 1989 (Expansión, various numbers). Employment in Mexico City grew only slightly throughout the 1980s, with increases in the service sector and a sharp decline in manufacturing (-94,000 between 1980 and 1989, or a reduction of 10%). Consequently, the share of the ZMCM in national employment in manufacturing fell from 45 to 33%. Mexico City's share also decreased in overall employment, although to a lesser extent (from 41 to 32%)(Aguilar 1996, cuadro 8.1, 8.2).
By the end of the 1980s, Mexico City's economy recovered from the crisis. Its share in the national GDP rose again, although without reaching the level of the 1970s. The recovery of Mexico City's economy was driven mainly by an upswing of the Federal District (as its decline in the 1980s was mainly caused by the capital's crisis). GDP in the Federal District grew by 3.5% annually (1988-1996), which is markedly above the national average (2.5%)5. Consequently, the share of the Federal District in the national GDP rose from 21 to 23% (1988-1996). The extent to which the city's economy recovered is shown also by the fact that the Federal District is one of the few Mexican states where GDP per capita was higher in 1995 than in 1980. While it decreased nationwide (-0.2% annually) and grew only slightly (0.2%) in the dynamic Northeastern border region (Coahuila, Nuevo León, Tamaulipas), in the capital it increased by 1.4%. Consequently, GDP per capita in the Federal District exceeded the national average in 1995 the by 3.3 times, while in 1980 it was only 2.6 times higher. That points to the concentration of highly productive economic activities (Pradilla Cobos 1997, cuadro 2; INEGI, various years [a]).
This recovery stems mainly from two trends. First, the manufacturing sector in the Federal District overcame the crisis regaining annual growth rates of nearly 3% (1993-1997). Consequently, the share of the Federal District's industrial GDP in the total Mexican manufacturing GDP remained constant in the 1990s on a level slightly more than 20%. Although this is, of course, much lower than in the 1970s, it is important to note that the downward trend in terms of production was stopped (INEGI, various years [a]). Thus, speaking of a "deindustralization" of Mexico City would be misleading for various reasons. First, due to the historical weight and the stabilization of the manufacturing sector, the ZMCM still is by far the most import single city for manufacturing in Mexico, although states like Tamaulipas, Aguascalientes, Chihuahua or Baja California have far higher growth rates in manufacturing. Second, employment, that was once formerly counted as manufacturing, might appear as part of the service sector due to internal reorganization of firms, although it still is attached to manufacturing. Third, part of the city's manufacturing might have become informalized, which would mean a reduction in statistically grasped employment, but not in output.
Economic recovery of Mexico City results also from the growth of the service sector, which became the most important sector both in terms of the GDP and in employment. Yet, the growth of the service sector does not simply reflect a shift towards services in general, since the sub-sectors "trade, restaurants and hotels" and "municipal, social and personal services" barley grew. Rather, the growth was concentrated on "transportation and communication" on the one hand, and "finance, insurance and real estate" on the other hand, with growth rates of about 3% annually (1993-1997). Consequently, the weight of advanced producer services within the Federal District's GDP grew from 11% in 1988 to 19% in 1997. Moreover, "finance, insurance and real estate" was the only sub-sector of the capital's economy that increased its share in the total national GDP (INEGI 1998; INEGI, various years [a]). That the economic upswing of Mexico City in the 1990s was very much linked to advanced producer services is further shown by the fact that employment in "real estate, financial and professional services" grew very fast - by 60% between 1990 and 1997. Amounting to 9% of the city's total formal employment, advanced services employed as much as half as many people as did manufacturing (INEGI, various years [b]). Finally, headquarters returned to the Federal District in the 1990s. In 1998, 213 of the Mexican "Top 500" were located in the Federal District (up from 145 in 1989), while the state of Nuevo León, where Monterrey is situated, had only 59 (1989: 51), and Jalisco, housing Guadalajara, only 40 (1989: 40)(Expansión, 21. Julio 1999).
During the process of recovery, two major socio-economic transformations happened to the city. First, its economic profile was changed. While until the 1970s trade and manufacturing were the most important sectors, by 1997 "municipal, social and personal services" dominated (31% of the capital's GDP), followed by "trade, restaurants and hotels" (22%) and "manufacturing" and "finance, insurance and real estate" (19% each)(Garza/Rivera 1994, 106-111; INEGI, various years [a]). Second, economic recovery was reached at the expense of employment and social standards. It is important to note that the economic improvement did not result in the creation a substantial number of new jobs in the formal economy. Though employment grew in the first half of the 1990s, it grew slower than the Mexican average, resulting in further decrease of Mexico City's share in national employment. In particular, manufacturing did not recover its capacity to generate employment. Industry, which provided more than 50% of formal employment in the ZMCM in 1980, employed only 20% of all occupied people in 1999 (Aguilar 1996, cuadro 8.1, 8.2; Cárdenas Solórzano 1999, 239).
Economic recovery was obtained not only at the expense of employment, but also income. Real minimum wages in Mexico lost, as indicated above, more than two thirds of their value in the last two decades. Yet, the decrease was even more pronounced in the Federal District. While in 1980 real minimum wages paid in the capital were about 20% higher than in the national average, this difference had almost disappeared by the second half of the 1990s (Boltvinik 1995, 37; Cárdenas Solórzano 1999, 246). Furthermore, evermore people work in the unregulated informal economy, where work tends to be poorly paid. Although by the very nature of the informal economy it is difficult to assess exactly how many people work there, some studies (of the International Labour Organization [ILO], for example) estimate that the number might reach to 50% or 60% of the economically active population - a sharp increase since the 1980s. Even the "National Institute for Statistics, Geography and Computer Sciences" (INEGI) indicates that 50% of the workers in the ZMCM are missing any welfare benefits, and 45% work in small establishments (1-5 workers), which are usually considered as informal enterprises. According to the INEGI, the number and the share of people who work under precarious conditions increased notably in the 1990s. Furthermore, it is interesting to note that various indicators suggest that informality is higher in Mexico City than in the national average. It is hence no wonder that a recent study concludes that (upward) social mobility nowadays is no more than an illusion (El Universal, 31.3.1997; Delgado Selley 2000; OECD 1999; INEGI 1999, 56; INEGI, various years [b]; Cárdenas Solórzano 1999, 243-245; Molina Ludy/Sánchez Saldaña 1997).
In regard to social polarization, data on the income structure of the employed population in the ZMCM (INEGI, various years [b]) have to analyzed carefully, since the INEGI does not reflect on the loss of the real value of wages. Yet this is a decisive factor, since a person earning five minimum wages in 1999 in reality obtained only slightly more of real purchasing power than a person earning two minimum wages in 1987. If this sharp contraction of real wages is taken into account, the notable decrease of people earning less than two minimum wages between 1987 and 1999 (from 83.6 to 56.1%) does not mean that the actual share of people earning incomes below the poverty line really declined. On the contrary: Taking five minimum wages as the income necessary to purchase the basic basket of goods at the end of the 1990s, compared to two minimum wages in the mid 1980s (La Jornada, 27.12.1996; El Financiero, 21.1.2000), the share of the employed population which obtained incomes below or up to the poverty line increased slightly (from 83.6 in 1987 to 85% in 1999). On the other extreme of the income scale the group with the highest earnings grew quickly, although it is still very small (amounting to 2.8% in 1993). Thus, in terms of earnings of the occupied population one can identify a trend towards polarization. However, data on the income distribution among households of Mexico City point to a somewhat declining inequality through the 1990s. The richest tenth obtained 34.6% of all income in 1996, compared to 39.5% in 1989. Conversely, the poorest 30% expanded their share from 8.9 to 10.1% (INEGI, various years [c]). However, this result has to be taken carefully since some distortion might result from the fact that city related data for the 1980s are missing. This decade was, at least on the national level, one of a pronounced growth of inequality (CONAPO 1999, 154).
According to data on the labor market, Aguilar (1996, cuadro 10) identified a growing polarization between jobs at the top and at the bottom of the hierarchy for the time between 1970 and 1990. This tendency seems to continue, although in an altered and less pronounced way. Between 1992 and 1998, petty traders and street vendors, a group that certainly corresponds to the bottom of the labor market hierarchy, expanded their share in the whole urban employment by 5%. The upper stratum (professionals, senior officials, senior management) grew also, though on a slower path, expanding its share in overall employment by 1.1%. The second fastest growing group was manufacturing workers (3.6%)(INEGI, various years [b]). It is not clear if blue collar workers should still be counted as lower middle class (as it was common until the 1980s). Yet, both the above cited data on the growth of precarious work conditions and the general trend to dismantle social protection according to the maxim of deregulation suggest that the growth of the group of manufacturing workers would add to the layers at the lower end of the labor market hierarchy.
It is important to note that in a time of impoverishment and growing inequality Mexico City is experiencing political opening. After decades under direct control of the President of the Republic, the nation's capital got an "Assembly of Representatives of the Federal District" in 1988, which was later transformed into the "Legislative Assembly of the Federal District". Citizens of the Federal District were admitted possibilities for formal participation by law ("Ley participación ciudadana", 1998), and governance is becoming somewhat decentralized. In July 1997 for the first time since 1928, the capital's citizens were able to directly elect their chief executive officer (formerly an appointee or "regent" of the president). The fact that the election was won by the candidate of the opposition (non-PRI) party PRD, Cuauhtémoc Cárdenas, was seen by many as an important step towards democratization of the city and the country (Ward 1999, 14; Ziccardi 1998). Yet, it still remains to be seen whether the fading away of patronage, corporativism and clientelism (which, although far away from being democratic, provided some sort of social reconciliation and political control) is followed by a system allowing for real participation and democratic control.
A final transformation worth mentioning is the turn-around regarding migration patterns. Mexico City, which has been the most important destination of migrants for decades, turned - in only one decade - to the country's main sending area. The migration balance of the ZMCM, that is, of the whole urban agglomeration, became negative in the second half of the 1980s, amounting to a net loss of 223,700 people6 (Parnreiter 1998a, 118-130). In other words; Mexico City grew not because of immigration, but despite emigration. The, compared to earlier decades, modest population growth was driven only by birth rates being higher than mortality. Regarding the 1990s, it is not possible to reach a definitive conclusion, because the census of 1995 (INEGI 1996) provides information on only immigration but not on emigration on the level of federal states. Furthermore, the data available is contradictory. However, most researchers (Porras Macías 1997, 43, 62-65; Gómez de León Cruces/Partida Bush 1996, 17) expect the trend towards emigration to continue, although maybe to a lesser extent than in the 1980s. Yet, the strong growth in internal migration in the 1990s and the fact that nearly 50% of the overall increase is reported in the State of Mexico (which houses half of Mexico City's population; see footnote 3) could probably point to a reinforcement of migration flows towards Mexico City in the 1990s (Parnreiter 1998a, 194-198). Nevertheless, between 1980 and 1995, the combined share of the State of Mexico and the Federal District (that is, in essence, the share of Mexico City) in all internal migrants shrank from 47 to 42.6%. Federal states attracting a growing number of migrants are Baja California and Chihuahua in the North, Morelos and Puebla in the center and Quintana Roo in the South-East. The emergence of new patterns of migration becomes even more evident if one excludes intra-urban mobility from data (see footnote 7) and focuses on migration balances instead of gross immigration. In that sense, between 1985 and 1990, the most important destination for internal migrants was Baja California, followed by Chihuahua and Quintana Roo. The State of Mexico came in only in fourth place, and the Federal District had, as already mentioned, a negative migration balance (Parnreiter 1998a, table A-11, A-14).
DEBT CRISIS AND THE END OF IMPORT SUBSTITUTION
The loss of Mexico City's share in national GDP and employment and the emergence of new migration patterns may, to a certain extent, be seen as decentralization of the country's economy and population. In fact, since the 1970s, Mexican governments had elaborated various plans to promote more balanced development. However, in the opinion of many researchers the de-facto decentralization that happened in the 1980s cannot primarily be attributed to these plans (Pradilla Cobos 1993, 39; Hiernaux Nicolas 1995, 158; for a different position see Ward 1998, 45). Some studies maintain that diseconomies of scale (e.g. high land prices, traffic jams or pollution) affected industries with high consumption of land, water and infrastructure, which might have forced them to relocate in nearby states in central Mexico or in the Northern border region (Bataillon 1992, 79; Davis 1993, 79f; La Jornada, 15.12.1996).
However, from my point of view, the most important reasons causing the erosion of Mexico City's position in the national economy in the 1980s are the definite end of import substitution and the crisis which accompanied the transition to another system ("globalization", "neoliberal modernization"). Data suggest that the first half of the 1980s was particularly critical to the ZMCM, and that this city was harder hit by the crisis than other cities. This may be surprising, given that Mexico City was the unchallenged "epicenter" of the Mexican economy with a rather developed and diversified economic base, above national average level of productivity and considerable foreign investment (Garza 1985). To understand the specific impact the crisis had on the ZMCM, it is necessary to focus on the particular quality of the crisis.
The immediate reason for the crisis of 1982 was that indebtedness was so high that Mexico could not pay interest anymore. However, this bankruptcy had deeper reasons than just a "simple" insolvency. Rather, 1982 marks the definitive end of import substitution and thus can be seen as a watershed in Mexican history. The point was not about getting money to become liquid again, but about the end of a specific development course pursued for five decades. Problems with import substitution began to show by the 1970s. Yet, thanks to the oil boom in the second half of that decade and the fact that foreign credit was easily available and rather cheap by that time7, the Mexican government could disregard the first signs of serious economic problems. However, terms of trade for oil worsened to the disadvantage of Mexico. When interest rates in the U.S. increased drastically in the early 1980s, Mexico could not meet its obligations (Arrighi 1994, 322-324; Dussel Peters 1997, 134, 147).
Though import substitution was pretty successful through the years, it never developed a base for intensive growth. Rather, it had to rely all the time on extensive growth, which means a steadily increase of inputs. In the 1970s, agriculture slid into a profound crisis, making it unable to generate export revenue in order to pay for manufacturing imports. Industry itself lost its capacity to achieve productivity gains; plants and technology became obsolete. Consequently, the Mexican economy could not generate the amount of exports and hence of foreign currency which would have been necessary to finance imports. Thus, inflation increased, problems in the balance of payments became notorious, and indebtedness rose drastically (Pradilla Cobos 1993, 15-25; Lustig 1994, 31-48; Dabat 1995, 870).
Why was it Mexico City to be hardest hit by the collapse of import substitution? Economic crisis, such as growth, is a phenomenon of social and spatial unevenness. The impacts of crisis do not affect all cities or regions to the same extent, nor are they evenly distributed among economic sectors. In general, branches hardest hit are construction, durable consumer goods and capital goods, whereas non-durable consumer goods fare better in times of recession. Uneven impacts of the crisis on different economic sectors have their spatial expression. Since nearly two thirds of industries producing durable consumer goods and capital goods were located in the ZMCM, it is no wonder that Mexico City was stronger affected than other Mexican cities (Garza 1985, 418-412; Garza/Rivera 1994, 11-14).
The re-orientation of production towards foreign markets further deteriorated Mexico City's position. With production gearing to exports, domestic markets lost importance, devaluing the large urban agglomerations first as markets and then as production sites. In the neoliberal model, the "masses" aren't needed as much as consumers as in the era of import substitution, and as workers they came under pressure because economic success depends ever more on rationalization and gains in productivity. Moreover, wage reductions that were crucial to become competitive in the world market reduced domestic purchasing power, additionally contributing to the devaluation of domestic markets. By far the biggest market, Mexico City was hardest hit by these developments (Connolly 1993, 66). Moreover, the strong impact of the crisis on Mexico City is explained by the fact, that precisely because it has been the epicenter of the Mexican economy, restructuring started there. New technologies and innovation in organizing production were introduced first in the ZMCM (Aguilar et al. 1996, 187), which resulted, at least in the first years, in high social and economic costs.
If Mexico City's problems in the 1980s result from the "debt crisis" and the strategies to overcome it, the question arises to what extent this crisis was of "national" or "global" character. In other words; can one relate the end of import substitution to global dynamics? Obviously, many countries had similar problems. In fact, import substitution as pursued in Latin America or in the Soviet Union collapsed practically everywhere within only a few years8 . More analytically, import substitution never was separated from the world market. It was only the severe crisis of world capitalism between the late 1920s and the end of World War II that opened the way for peripheral states to implement import-substitution strategies. Though for many states these were quite successful (e.g. the production of consumer goods), when it came to capital goods, which demanded a more sophisticated technology, it became evident that dependency on imports of technology and capital could not be overthrown. Indebtedness grew, with well-known consequences. "First World" bankers also created indebtedness. Because of a structural crisis of the world system, investment in industries in core countries became less profitable in the 1970s. One consequence was that overabundant capital was channeled as credit to "Third World" countries, which for them turned out to be a boomerang only a few years later. In sum; the argument can be launched that the breakdown of import substitution was a worldwide phenomenon, closely linked to the structural crisis of the capitalist world system. This crisis caused the end of the specific fordist arrangements as known from the 1950s to the 1970s in the centers, and it caused the end of "incomplete fordism" or import substitution in the peripheries (see, for example, Amin et al. 1982; Frank 1990, 26; Arrighi 1994, 322-324; Altvater/Mahnkopf 1996, 405-409).
ECONOMIC RECOVERY, RE-CONCENTRATION AND GLOBAL CITY FUNCTIONS
The 1980s were, without doubt, a period of (unintentional) decentralization, which can be attributed to the specific spatial expression of the crisis and of the shift towards a new economic strategy. Decentralization happened, on the one hand, towards the Northern border region, and, on the other hand, to the central region, particularly to the so-called "crown cities" of Mexico City Toluca, Cuernavaca, Pachuca, Puebla y Tlaxcala. The key question for the 1990s and the future is how restructuring and re-orientation of the economy have affected and still affect Mexico City. After 1982, there was only one imperative for Mexico - to restore the ability to make interest payments. Policies applied to achieve this aim were sketched out above. The focus here lies in the spatial impacts of neoliberal modernization and globalization, and in particular, on its impact on Mexico City. Some argue that in an open economy a trend towards decentralization is sustained, because principal markets are located abroad, which reduces the advantages for firms to be located in the main urban agglomerations. Given the relatively high costs of land, labor, transportation etc. in Mexico City, companies would tend to locate in cheaper areas or/and closer to their markets (Livas Elizondo 1994; Connolly 1997). However, others maintain that globalization increases competition, and thus forces companies to rely more on scale production. That would imply the concentration of production, infrastructure etc. in a limited number of places, of which Mexico City would be the most important. Likewise, the growing importance of finance reinforces centralization even more (Hiernaux Nicolas 1994; Rivera 1997; Pradilla Cobos 1997).
Although until now data do not allow for a definitive conclusion, there is some support for the second hypothesis9 . As already mentioned, in the late 1980s and in the 1990s Mexico City's GDP grew again, and so did its share in the national GDP. This upswing was driven mainly by the recovery of economic dynamic in the Federal District, which is indicated also by the return of many headquarters to the capital. Both the rising GDP and the return of headquarters show that the ZMCM had begun recovering from the severe crisis and that, at least to some extent, re-centralization is occurring. However, in the context of global city research, neither GDP nor the location of headquarters in general are decisive factors.
The task then is to link the new dynamic of Mexico City's economy to new functions, namely those which serve the globalization of Mexico and which connect the city to other global cities. Documentation that such ties do exist is still tentative due to severe data problems which in general challenge the global city research (Short et al. 1996, Taylor 1997, 1999a). Most data available are state-centric, and if data are on cities, they commonly are highly aggregated. Furthermore, the basic unit usually covered by city data is the city as an entity and not as an element in a "space of flows", which allows only for indirect conclusions about relations between cities and hence the global urban system. Finally, an analysis of Mexico City as part of a global, cross-border network of cities faces another problem. Since the 1980s, we can observe the formation of a very dense and complex regional division of labor in the center of Mexico, which connects Mexico City closely and functionally to six other federal states. Whether this regional system is called "megalopolis del valle de México" (megalopolis of the Mexican valley), "región centro" (central region) or "corona de ciudades" (crown of cities), the point is that one cannot separate Mexico City from its regional embeddedness. However, data on this broader spatial unit do barely exist. Despite all these data problems some evidence for the formation of a global city is provided in the following sections. In addition, I also will sketch out points for a research agenda.
Location Patterns of Headquarter
A first step to study the relationship between the new dynamic of Mexico City's economy and global functions is a closer analysis of the location patterns of the most important firms registered in Mexico. The results (see Table 1) can be summarized as follows: the bigger the firm (in terms of sales) and the stronger the links to the global economy (in terms of exports and imports and of participation of foreign capital), the higher is the probability that a company has its headquarter in the Federal District. All other Mexican states, including Nuevo León (Monterrey), are far behind.
Regarding the sales, 70% of the "Top 10" Mexican companies were headquartered in the Federal District in 1998, compared to 42.6% of the "Top 500". The dominance of Mexico City becomes even clearer if one adds those communities of the State of Mexico to the Federal District, which belong to the ZMCM. In that case, half of the biggest 500 companies have their main office in Mexico City. However, it is important to note that the high concentration of headquarters is not translated into a concentration of production and employment in Mexico City, since most of the big firms have spread their production to various cities in the whole country (see below). The only other Mexican state attracting a significant number of main offices is Nuevo León with its capital Monterrey. Similarly to Mexico City, the percentage of companies being headquartered in Nuevo León grows with the total of sales.
Since the primary concern of this article is the global embeddedness of Mexico City, the most meaningful information that can be drawn from location patterns of headquarters is data that specify global links. Participation of foreign capital is one indicator for global links. Results reveal that companies which are dominated by international capital show a much stronger preference to locate in Mexico City than nationally owned (private) firms. As indicated above, 42.6% of all companies recorded by the ranking of the "Top 500" are headquartered in the Federal District. Yet, if the list is broken up according to the ownership of the firms, one easily distinguishes different location strategies between nationally owned (private) companies and those with a majority of foreign capital. Among the firms registered by the list of the "Top 500" Mexican companies which are dominated by private Mexican capital, 39.4% have their main office in the Federal District. By contrast, among the foreign dominated corporations listed in the ranking, 58.3% have their headquarter in the Federal District. Again, concentration even increases with the size of the firm and if one considers the whole metropolitan area (ZMCM). For example, 78,9% of companies that are among the biggest 100 firms and that are internationally dominated are headquartered in Mexico City. Again, it must be taken into account that these data do not refer to the whole production of a firm, but rather to a specific segment of it, namely control functions and services.
Regarding foreign trade (except the maquiladora industry10 ), which is another important link to the global economy, the results show the same tendency as in the cases of sales and participation of foreign capital: there is a very strong dominance of Mexico City as preferred location for headquarters, and the preference increases the more a company exports or imports. Nine out of the ten biggest export firms are located in the Federal District, compared to 43% of the 300 firms with most exports (ZMCM: 51.3%). Nuevo León comes next, but with a large backlog. A similar pattern is found in regard of imports. The dominance of Mexico City as location for main offices of export firms is shown also by the fact that 73.4% of the exports of the biggest 100 export companies (again without maquiladora industry) originate in the Federal District, a huge lead over Nuevo León (11.5%). Remember, that this concentration of export companies in Mexico City does not mean that commodities for export are actually produced there. Rather, it points to the city's role as central node in the international and regional division of labor (see below).
Yet, one should know the nature of export firms based in the Federal District. A clue in this regard is productivity, which points to global competitiveness. As already mentioned, GDP per capita in the Federal was considerably more over the national average in 1995 than in 1980 (INEGI, various years [a]). This gain in productivity stems, at least partly, from the manufacturing sector. In an analysis of the Mexican industry (without maquiladoras), Dussel Peters (1997) identifies automobiles, petrochemicals, electrical equipment, beer and glass as the sub-sectors with the highest gains in productivity. A majority of the most important firms of these branches have their headquarters in Mexico City (the exceptions are the beer industry, which shows a tendency to locate main offices in the Northern states, and the glass industry, which is comparatively small). In the Federal District are located 4 out of the 7 automobile firms recorded among the "Top 500" Mexican companies, 9 out of the 14 listed firms in the petrochemical industry (and another one lies in a community of the State of Mexico belonging to the ZMCM), and 5 out of the 12 companies in electrical equipment (3 more are located in communities of the State of Mexico belonging to the ZMCM). Thus, two thirds of the headquarters in these three very productive branches are located in Mexico City. This is a considerable higher share than among all Mexican firms, what allows for the postulate that productivity grows faster among firms based in Mexico City than in the Mexican mean. Once again, these data do not necessarily refer to the actual production of automobiles or electrical equipment. Rather, they indicate wherefrom strategies to raise productivity are developed or at least implemented.
To sum up; the Federal District is the preferred location for headquarters of the biggest Mexican companies. This preference increases with a) the number of sales of a company, b) a higher level of participation of foreign capital, c) a stronger orientation towards export-production, and d) higher levels of productivity. This specific character of main offices in the Federal District points to a close link of the city's economy to the global economy. Data suggest that companies which were able to adapt to the world market show a preference to locate in the Federal District. In other words; Mexico City is the place where the "global player" among Mexican companies - and transnational companies settled in Mexico - tend to have their headquarters. Consequently, the Federal District assumes world city functions in the sense that from there the Mexican economy (or at least parts of it) become globalized. If that is true, Mexico City's loss in manufacturing employment represents more than a simple "deindustralization". Rather, "deindustralization" might fit into a new regional and national division of labor, in which Mexico City specializes primarily on functions commonly attributed to global cities, while other cities in the central region and in the North take on manufacturing. However, detailed documentation about functions carried out by these headquarters and the networks linking them to production sites elsewhere in Mexico and foreign markets are still lacking.
Taking up the question whether globalization would favor a more decentralized economy or not, evidence suggests that in fact re-centralization is happening. Further, concentration of decision making and high level economic management in the country's capital is particular pronounced among companies with strong linkages to external markets. To explain this marked difference in location strategies, two hypothesis can be raised. First, the obviously greater preference of companies engaged stronger in the world market to settle in the country's capital might result from their wish to be close to the center of national political power. For example, foreign dominated companies might feel a greater need than national firms to be in touch with high government officials or high representatives of the country's main political party PRI in order to establish or expand business in Mexico. Second, location preference for the Federal District might also result from the need for highly specialized producer services which can be obtained only there. While a national firm might get by more easily on the services offered in a regional metropolis (e.g. Monterrey), a company catering to the world market might need, for example, expertise on U.S. law. This specific service, however, might be available only in the Federal District. Yet, the two presumptions are, of course, not mutually exclusive. Further research will probably show that it is a combination of political considerations and the need for highly specialized services which drives foreign companies to the country's capital.
Another task for further research is to map the network between headquarters in Mexico City and production sites elsewhere in Mexico. The Mexican branches of General Motors, Daimler-Chrysler, Volkswagen, Ford and Nissan are, for example, leading enterprises in terms of sales, exports and foreign investment11 . All of them have their headquarters in the Federal District, except Volkswagen, which is located in Puebla. Yet, production facilities are spread over the country. General Motors produces in Toluca (State of Mexico), Ramos Arizpe (Coahuila) and Silao (Guanajuato); Daimler-Chrysler in Saltillo (Coahuila), Toluca and in the Federal District; Ford in Hermosillo (Sonora), Chihuahua and in Cuautitlán, a community of the State of Mexico which belongs to the ZMCM; and Nissan in Aguascalientes, Cuernavaca and Zivac (both in Morelos), and Lerma (State of Mexico). In addition, Nissan has research and development centers in Manzanillo (Colima), Toluca and in Mexico City. Thus, though Mexico City is not totally negligible as manufacture site, evidence show that most of the automobile production takes place either in states in the country's center (State of Mexico, Guanajuato, Aguascalientes and Morelos), or in Northern border states (Coahuila, Chihuahua). The same applies to the computer industry, a fast growing branch with massive exportations to the United States. IBM, for example, is as well one of the biggest export firms in Mexico as an important recipient of FDI. Mexico is for that transnational company an ever more important production place - 55% of all laptops produced by IBM are assembled in Mexico. Yet, the company, which came to Mexico City in 1957, shifted its plants to the state of Jalisco in 1975, leaving in the Federal District only its main office (Dussel Peters 1999, 58-67). Further research should aim to analyze the links between the production sites, the (regional) headquarters in Mexico City and foreign cities, where the capital, specific services or semi-finished products might come from. That would allow for the conceptualization of the network or commodity chain in which main offices based in the Federal District articulate Mexican production into the world economy.
Regional Distribution of Foreign Investment
Another example of the growing embeddedness of Mexico City's economy in global dynamics can be drawn from capital flows. These represent, without doubt, one of the most important links between a "national" and the "global" economy. As already mentioned, an unprecedented inflow of capital is one of the most significant outcomes of the opening of Mexico's economy. Between 1989 and 1998, more 80 billion dollars were invested in Mexico as Foreign Direct Investment (FDI), three fourth of it after 1994. Portfolio investment, which outnumbered FDI in the first half of the 1990s, fell behind in recent years (1994-1998: 18.7 billion dollars), mainly due to the "peso crisis" in 1994-95 (Dussel Peters 1999, cuadro 1, 2; Consulado General 2000).
Both flows of capital are highly concentrated on the Federal District. Regarding FDI12 , 58.6% of all investment (that is, 27 billion dollars) effected between 1994 and 1998 was realized in the Federal District (see Table 2). Although through the last years a decrease in the share of the capital can be observed, the Federal District's participation is in fact the same than in previous years (1989-1993: 59%). It is noteworthy that the six Northern border states together (Baja California, Sonora, Chihuahua, Coahuila, Nuevo León, Tamaulipas), which concentrate 86% of all maquiladora production (Bendesky 2000), attracted less than half of FDI than the capital (28.4%)(own calculations, based on Dussel Peters 1999, cuadros 8, 9). The dominance of Mexico City becomes even more striking if one considers that most of the FDI realized in the State of Mexico might have been directed to firms located in communities which belong to the ZMCM. Thus, the share of Mexico City would nearly reach two thirds of all Foreign Direct Investment. For portfolio investment data on regional distribution is lacking. However, it is likely that the concentration on the Federal District is even higher than in the case of FDI, given that the stock market is located there, and all major banks have their main office - and hence their accountancy - in the capital. The weight of Mexico City as a financial center and therefore as a key place for globalization is underlined further by the fact that the Mexican stock market is the second largest in Latin America, and Mexico comes third behind China and Brazil among emerging markets in receiving foreign capital.
Yet, not all money that enters the Federal District is necessarily invested there (or, at least, not necessarily exclusively invested there). It might be, for example, that a transnational corporation which effects a "greenfield investment" sets up the new plant anywhere in Mexico, while the regional headquarter of that company from where "local" production is controlled and serviced is located in Mexico City. In 1998-99, for example, the biggest single investments were effected by Daimler-Chrysler and Ford (1.5 billion dollars each), both headquartered in the Federal District. Yet, the investment was directed towards the expansion of an existing plant in Coahuila in the case of Daimler-Chrysler and to the construction of a new one in Chihuahua in the case of Ford. The same principle applies to mergers and acquisitions. The U.S. American company Anheuser-Busch, for example, invested 556 million dollars to buy 13% of the Mexican group "Grupo Modelo". This worldwide known beer company ("Corona") has its main office in the Federal District and its breweries in eight cities throughout the country (one of them in Mexico City)(Dussel Peters 1999, cuadro 11; CEPAL 2000, 77). In these cases foreign investment is entered in the Federal District - because accountancy is carried out there - and then eventually been re-distributed to distant production sites in Mexico. Alike in the case of the headquarters, it is necessary to trace back the flows of capital - from the global investor to the local use. That would allow for conceptualizing an urban network, in which different cities occupy different positions with different tasks and power. To identify the "command and control" linkages between headquarters and specific services based in the Federal District and production facilities elsewhere in the country remains a key subject for research to sketch out the spatial and sectoral distribution of investment and the role which banks, financial institutions and other services located in Mexico City play in that division of labor.
Employment in the Advanced Producer Service Sector
Until now I have stressed the high concentration of headquarters (namely those of global oriented companies) and of Foreign Direct Investment in Mexico City. However, both headquarters and FDI are related to certain activities - in particular, to specific services. The Mexican branch of a transnational automobile company, a foreign investor buying assets from a former state owned enterprise, a Mexican beer company catering to the U.S. market or a broker speculating on currencies - they all need to be serviced by accountants, tax and financial advisers, lawyers, advertising agencies, political consultants etc. In other words: The work of integrating parts of the Mexican economy into the global economy requires specific services. Yet, global city researchers emphasize not only that advanced producer services are crucial because they provide the means by which globalization is done, but also that much of these services (particularly the ones which are not routinized) concentrate necessarily in major cities, because it is there where both the creative milieu needed to develop them and the markets to trade them are to be found (see, for example, Sassen 1991). Accordingly, the advanced service sector in Mexico should have, on the one hand, grown and, on the other hand, strengthened the position of Mexico City.
Indeed, both happens. As in other countries, the advanced producer services grew substantially in the last two decades. Its share in the Mexican GDP doubled between 1985 and 1993 to 16%, decreased notably in 1995 due to the "peso-crisis" and recovered since then. National employment13 in "real estate, financial and professional services" grew by 45% between 1992 and 1997 (OECD 1995, 176; INEGI, various years [a]; INEGI, various years [b]). This growth was strongly concentrated on Mexico City. With 563,724 people working in that sector in 1997, the ZMCM had 47.8% of all national employment in "real estate, financial and professional services", compared to Guadalajara and Monterrey, which had 9.3 and 8.7%, respectively. Although employment in the specialized service sector grew faster in these two cities in recent years than it did in Mexico City, and although Mexico City's share in the national employment decreased slightly as a consequence, the ZMCM's dominance is still obvious. Additionally, advanced services make a higher share of urban employment in the ZMCM than in any other Mexican city, exercising thus a stronger impact on the city's economic and social development. In 1997, advanced services amounted to 8.6% of all formal employment in Mexico City, or to 44% of the city's employment in manufacturing (INEGI, various years [b], see Table 3).
The (slight) decrease in the ZMCM's share of overall employment in "real estate, financial and professional services" does not necessarily mean that the Federal District's "command and control" capacity had diminished. The relative reduction in employment in this sector might well be the result of the decentralization of routinized tasks to other cities. In fact, this argument is supported by a growing share of the Federal District in the nation's GDP in "Financial Services, Insurance, and Real Estate". Although this growth was slight (1993: 26.5%; 1997: 27.4%), it is significant. It indicates, together with the shrinking percentage in employment, the higher productivity of advanced services in Mexico City than elsewhere. In fact, Mexico City is the only Mexican city where productivity in advanced services is above national average in all sub-branches, while rival cities such as Monterrey or Guadalajara exceed the national average only in half of the cases or even less (INEGI, various years [a]; Garza/Rivera 1994, 67, 73f, 90f, 106-110). The boom of the advanced service sector indicates that the new dynamic of Mexico City's economy in the 1990s is closely linked to functions related to economic globalization. This re-direction of the economy from activities orientated mainly towards the domestic market to external linkages changed both the economic profile of Mexico City and the city's social structure. The ZMCM turned from a city specialized in manufacturing, as it was until the late 1970s, into a city specialized in advanced services, and that both in terms of GDP and employment. According to the "Coefficient of Local Specialization", the Federal District was, in 1993, highly specialized in services related to finance and insurance, and, to a lesser degree, in professional services (Garza/Rivera 1994, 67, 73f, 90f, 106-110; Iracheta Carroll 1999, 118f).
Therefore, it is very likely that the recovery of Mexico City's economy may be, at least partly, due to the growth and concentration of advanced producer services (recovery is also based, as already mentioned, in a stabilization of manufacturing). In addition, in the case of advanced services, the new economic strategy certainly has given rise to re-centralization of economic power in the country's main metropolis. Finally, the high concentration of advanced services in the ZMCM shows that Mexico City, particularly the Federal District, engage ever more in global activities. That means that the articulation of export manufacturing (excluding maquiladora industry) or financial markets with the world economy, are mainly organized and controlled from Mexico City. In other words; the globalization of Mexico is produced to a high extent in Mexico City. All this is in line with global city research. However, for a more detailed study of the role of Mexico City in the cross-border urban network one should be able to disaggregate data. "Real estate, financial and professional services" is a fairly wide field, including a broad range of professions. That does not allow for distinguish specific tasks fulfilled in Mexico City. Additionally, one should know more about the activities and spatial distribution of firms that are serviced by Mexico City's advanced service sector.
To conclude: the analysis of location patterns of headquarters, of the regional distribution of Foreign Direct Investment, and of advanced producer services indicate that the opening of the Mexican economy and the stronger orientation towards global markets enforced concentration in Mexico City. This new centrality is particular seen in those activities that are essential to the globalization of the Mexican economy. Mexico City, and in particular, the Federal District, function ever more as a hinge between "Mexico" and the "world market", and in that sense one can speak of the making of a global city. In addition, the recovery from the economic crisis of the 1980s stems, at least partly, from this new role the city assumes. However, it is important to note that the economic upswing is not accompanied by a social recovery.
IMPOVERISHMENT, SOCIAL POLARIZATION AND NEW PATTERNS OF MIGRATION
The traditional factory worker, who lost his job during the crisis of the 1980s, is very likely to be unable to find employment in the advanced service sector. Rather, he might be forced to emigrate to the U.S. or to work in Mexico City's growing informal economy to make a living (de la Peña et al. 1990). Thus, the new role assumed by Mexico City effects not only its economic structure, but also social arrangements. A critical issue is the fact that economic recovery in the late 1980s and in the 1990s happened at the expense of most of the city's inhabitants. As pointed out above, the last two decades were a time of absolute impoverishment of large segments of the urban population, and they were also a time of growing polarization of earnings and labor market structure.
The arising question is whether impoverishment and polarization are linked to the deeper integration of Mexico into the world economy and the growing importance of global city functions within the economy of Mexico City. Evidence presented so far suggest that that is in fact the case. First, it is obvious - and admitted by high representatives of the government - that substantial wage reductions were a crucial element in the country's course of neoliberal modernization and globalization, since to cheapen production allowed to become more competitive both at the world and the domestic market. Furthermore, wage reductions were an important cornerstone in the efforts to reduce inflation. Second, the recovery of the manufacturing sector was succeeded, at least partly, by reducing employment and/or by downgrading production to the informal sector. As shown above, industrial employment stagnated although output grew in the 1990s, and work in the informal economy expanded at the expense of formal capital-labor relations. Third, the growing importance of services contributed both to impoverishment and polarization. In the service sector, more people than in the national average gained wages under the poverty line14 (72% compared to 70%), and more people were really well off, earning more than ten minimum wages (3% compared to 2.8%). In contrast, manufacturing, which also had an above average number of poorly paid workers, was underrepresented among the highly income groups, too. The majority of the well off people in services worked in the sub-sector of the advanced producer services, making this branch to the one with the highest number of people earning more than ten minimum wages (note that the manufacturing sector, which employed three times as many people as "real estate and financial and professional services", had less high income employed). The polarizing structure of the labor market point into the same direction (INEGI, various years [b]). Yet, the point is not that globalization created the informal sector or income inequalities in Mexico City. Rather, by strengthening links to the world economy, Mexico City experiences the worsening of social problems it already had before. This is concurrent with the findings of world city researchers, who argue that a growing informalization of the relations between capital and labor and a deepening of social polarization are characteristic processes of world city formation.
While social polarization and in particular its relationship to globalization needs to be studied further, in the case of the re-orientation of migration flows we have solid evidence for the impacts of globalization (for more details, see Parnreiter xxxx). The emergence of new patterns of internal migration in Mexico, in which Mexico City is an area of net emigration and in which the northern states of Baja California and Chihuahua and the southern state of Quintana Roo became main destinations for migrants, is closely linked to the altered position of Mexico in the national and global economy (Corona/Tuirán 1994, 22). First, since the economic crisis reduced possibilities for employment and earning in Mexico City, people may have been motivated to leave the city for other places. The Federal District15 recently has had a negative migration balance with most of the traditional sending areas such as the federal states of Puebla, the State of Mexico16 , Veracruz or Hidalgo. Another option for people who wanted to leave the capital was the removal to nearby states such as Morelos, although in such cases people normally continued to work in the Federal District. Finally, emigration from Mexico City to economically more promising cities and to the United States became an phenomenon of increasing importance. It is mainly the northern state of Baja California that has rapidly gained importance as a destination for emigrants from Mexico City (INEGI 1995, 12ff, 137; Corona Cuapio/Luque González 1992, 27).
The second consequence resulting from the economic crisis is that Mexico City lost attractiveness for internal migrants. Between 1980 and 1985, the ZMCM attracted 76,589 migrants less than ten years before, which means a reduction of more than 15% (Corona Cuapio/Luque González 1992, 26; INEGI 1994, 50). Taking into account the growth of immigration in the new industrial cities in the North and in the centers of tourism in the South, this reduction very likely is due to the greatly reduced capacity of Mexico City to create employment. The demand for labor, which drew millions of migrants to Mexico City since the 1930s, did not serve as an engine of immigration anymore. In the 1980s, new industrial workers were needed mainly in the North of Mexico, and the workforce serving international tourism was needed primarily in the South.
MEXICO CITY IN THE CROSS-BORDER NETWORK OF CITIES
Although still tentative, evidence presented so far supports the argument raised by world city research that globalization produces a new geography of centrality in which major cities are nodal points. This seems to hold true for Mexico, too. Evidence suggests that a specific set of structural transformations going on in Mexico City is linked to globalization and to the particular role of Mexico City in this process. Parts of the city engage increasingly in global activities, so that Mexico City, and in particular, the Federal District become a hinge between "Mexico" and the "world market". In that sense we can speak of the making of a global city.
Support for this argument comes from various studies of the "Globalization and World Cities Study Group and Network" (see Beaverstock et al. 1999a, 1999b, 1999c; Taylor/Walker 1999; Taylor 2000; Taylor et al. in this volume). Their review of 15 of the most relevant publications on world cities shows that Mexico City is listed by six studies as a world city, ranking 21st among all cities cited. In order to achieve a refined hierarchy, the "Globalization and World Cities Study Group and Network" analyzed the location patterns of 69 firms in four key service sectors (accountancy, advertising, banking/finance, and law) among 263 cities in 1997-98. They found that 122 cities served as the location for at least two firms in at least one of the sectors. By evaluating levels of presence of firms in each branch on the basis of size and importance of offices (from 0 for not qualified to 3 for prime center), a ranking was produced, with a maximum of 12 scores and a minimum of 1. Cities with at least 4 scores, or, in other words, which are global service centers at least in two branches, are identified as world cities - 55 altogether. Mexico City17 ranks in the 15th place in this hierarchy, with 8 scores identified as a "Beta World City". It shares this position with Brussels, Madrid and São Paulo, only narrowly behind San Francisco, Sydney, Toronto and Zurich (with 9 scores each), but ahead of Washington, Miami, Berlin or Shanghai. In an improved measurement, Mexico City is listed in the 20th place -as the highest ranked Latin American city. It has 12% of world city formation (measured in terms of the level of service provision of the mentioned sectors relative to the top scoring city), ahead of São Paulo (with 11% of world city formation), Buenos Aires (6%), Caracas (6%) and Santiago (5%). Mexico City can be compared to cities like Zurich (11%), Johannesburg, Milan (13% each) and even Los Angeles (14%), which is an surprisingly high classification, given skeptical judgments such as; "Mexico City's future as a world city is far from clear" (Friedmann 1995, 38).
The high ranking of Mexico City is based on the strong presence of "global players" among service firms. Mexico City qualifies as a "major" global service center for all four sectors (accountancy, advertising, banking and legal services)18 , while São Paulo is a "major" global service center in three sectors (advertising, banking and legal services) and Buenos Aires only in one (banking). The fact that Mexico City houses many of the global service firms suggests that the Mexican capital is well embedded into the cross-border network of cities. Indeed, further evidence obtained by the research of the "Globalization and World Cities Study Group and Network" confirms this assumption. Firms in the advanced producer service sector that are located in London but operate on a global scale tend to have a high presence in Mexico City. In finance and banking, for example, there is a 93% probability that a global firm located in London has a direct link to another of the firm's branch in Mexico City. Considering that only two cities show a higher probability of direct organizational links to London (New York and Tokyo) and taking into account that London based finance firms are as present as in Mexico City only in three cities (Buenos Aires, Hong Kong, Singapore), Mexico City can in fact be considered as a strategic location for globalization processes. In advertising the organizational links to London are slightly weaker - the likelihood that a London located global firm has a direct connection to another of the firm's branch in Mexico City amounts to 82%. In law the ties are still very weak - the probability for a London located law firm to have direct link to Mexico City is only 5%19.
Addressing not only presence of London based firms, but the broader question of connections of cities to London by averaging across the four producer services (accountancy, advertising, banking/finance, and law), Mexico City comes in 15th place among 54 cities. Of course, the "Alpha World Cities" New York, Paris, and Tokyo (in that order) have stronger links to London, as do major European cities (such as Brussels, Frankfurt, Milan or Madrid) and the Asian world cities Hong Kong and Singapore. Mexico City's connectivity to London is the same as São Paulo's and notably stronger than that of Buenos Aires. It compares to the linkages of Chicago, San Francisco, Washington D.C., Dusseldorf or Amsterdam. This is, again, a surprising result, underlining Mexico City's importance as a strategic site in globalization. The significance of having strong linkages to London even gains weight when one considers that Mexico City a) has no historical ties to London (as Hong Kong or Sydney), and b) forms part of another regional bloc in the world economy (and hence is supposed to be more closely connected to U.S. cities).
Analyzing geographical strategies of U.S. law firms in 1997, the "Globalization and World Cities Study Group and Network" again identifies Mexico City as a strategic place in the cross-border network of cities. Eight of the 368 foreign offices of U.S. law firms are located there, which represents 2.2% of all. This seems to be a rather negligible number. However, a closer look reveals that though the number is small, Mexico City still classifies high in location strategies of U.S. law firms. First, one has to take into account that the number of U.S. law firms in Latin America is unexpected low - 23 altogether, suggesting that this region cannot be considered as a major globalization arena. Second, out of these 23 offices, Mexico City houses eight. Thus, it is the only Latin American city with a significant presence of U.S. law firms. Third, Mexico City is throughout the world the most important location for U.S. law firms outside the three areas identified by Beaverstock et al. (1999a) as globalization arenas (Western Europe, Pacific Asia, Eastern Europe). Consequently, despite having only 2,2% of all foreign offices of U.S. law firms, Mexico City still ranks 11th in their location strategies.
The studies of geographical strategies of firms do not provide data on the actual transactions of companies. Additionally, the analysis of the relations of London based service firms to other cities includes firms that might have their main office elsewhere in the world. For example, the three U.S. law firms with the most pronounced global strategies (Baker & McKenzie, White & Case, Coudert Brothers) appear in the list of London based global firms, although they are from Chicago and New York, respectively. This means, for the purpose of this article, that though we can conclude from the data cited that Mexico City is a rather prominent place in the geographical strategies of global service firms, we cannot guess where an international investor buys the services to conduct transactions. It might be in London, but it seems more probable that he would turn to the New York or Chicago headquarter or branch. Thus, the actual space of flows created by services remains as well a subject for further investigations as the place of Mexico City in it.
Connections between cities need to be built up. Inter-urban telecommunications networks are a key infrastructure to link distant financial markets, firms producing advanced services and off-shore sites for manufacturing. Faxes, telephones and internet connections are necessary to integrate and mediate the globally stretched net of financial and other business activities. The emerging global telecommunications system creates, similarly to the net of global service firms, an uneven geography, because technical facilities are by no means evenly distributed. On the global scale telecommunications networks contain "hubs" and "spokes", and on the city scale infrastructure and equipment are highly concentrated, too. Thus, the Central Business Districts of global cities are the very centers of the inter-urban telecommunications network, because they dominate both investment in and use of complex information and communication technologies. Centralization of the advanced telecommunications infrastructure is promoted further by the fact that the economic activities which rely strongest on them (such as finance) are precisely the ones which have the greatest need for social connectivity and face to face interactions (Graham 1999; Sassen in this volume).
The availability of ever more sophisticated, diverse, capable and cheap telecommunications technologies has become a key asset in the competition between cities to attract financial and corporate operations. Concurrent with the above quoted findings of Taylor and his colleagues, a survey which classifies the competitiveness of telecommunications provision in July 1998 in 25 major cities20 reveals that Mexico City is well integrated into the inter-urban telecommunications networks. Mexico City ranks 16th - better than any other "Third World" city21 and better even than Zurich or Singapore. While the pricing of services is still one of the biggest disadvantages of Mexico City, the availability of sophisticated connections and the choice of physical infrastructure connections accessible were appraised particularly (Finnie 1998, cited in Graham 1999, 938).
The relative high ranking of Mexico City results from efforts undertaken in the last decade to build up an adequate infrastructure to compete in the world market. Privatization of the state-owned telephone company (Teléfonos de México) in 1990 was a first key step in this direction. The Mexican Grupo Carso and two major "global players" among telecommunications companies took control over Telmex - Southwestern Bell and France Telecom. The latter two groups, despite having only 10% of the capital of Telmex, control 49% of the stocks. They were crucial to expand and, in particular, to modernize telecommunications in Mexico inasmuch as they provided latest technologies. In the words of the Federal Commission of Telecommunication: "The integration of these companies into Telmex guarantees the development of the most modern telecommunications network, promoting economic progress in Mexico. Alike, it opens the door to a technological revolution" (COFETEL 1998a, my translation).
In fact, since the Mexican telephone company was "efficiently colonized by foreign owned telecoms operators" (Graham 1999, 939), facilities needed for secure, fast and voluminous data transmission and hence for cross-border communication expanded and improved rapidly. The number of telephone lines doubled to more than ten million between 1990 and 1999, and the quality of lines - an equal or even more important factor for efficient data transmission - improved substantially. In 1999, more than 97% of all lines were digitalized - up from 29% in 1990. In 1994, an electronic backbone consisting of two 2 megabits fiber optic lines and several 64 kilobit lines was set up (and expanded since then), connecting the main Mexican cities to the United States. This infrastructure helped to expand conventional telecommunications like faxes or telephone calls (international calls, for example, grew by 436% between 1990 and 1998 up to 737 million), and they allowed Mexico to join the Internet. Since the first Internet connection in Mexico was put into place in 1989, both infrastructure for the use and internet communication itself expanded rapidly. In July 1999, the number of Mexican hosts22 totaled 224,239 - up from 13,787 in January 1996. Mexico ranks second among Latin American countries in terms of both growth rates and the absolute number of hosts - surpassed only by Brazil -, and it ranks 20th worldwide. Alike, the number of Internet users increased from 94,000 in 1995 to more than 1.3 million in 1998. The development of the infrastructure was driven mainly by business interests. While the first Internet users were universities, by now business related users represent the majority - and their dominance grows (56% in 1998, up from 51% in 1995)(COFETEL 1998b, 1998c, 1999a, 1999b, 1999c; Fernández 1995; Gutiérrez/Daltabuit 1999, 23; ISC 1999).
These data suggest a fast growing integration of Mexico into the cross-border information flows. However, it is not the whole Mexico which is part of the global telecommunications network. The upgrading of the Mexican telephone system and the new facilities for telecommunications reinforced an uneven geography of information flows. For example, more than 38% of all Mexican telephone lines are concentrated in Mexico City, which is twice as much as the city's share in the country's total population, and nearly a third of all international telephone calls are made or received from Mexico City. More significant is, however, that the Federal District is the central node of the national electronic backbone, which now consists of various 2 megabits fiber optic lines linking 18 cities. Five lines meet in the Federal District: the South-East connection to Mérida, crossing Puebla, Veracruz, Villahermosa and Campeche; the North-East connection to Monterrey and Matamoros; a single connection to Querétaro in the central region; the North connection to Ciudad Juárez crossing Celaya, León, Aguascalientes, Torreón and Chihuahua; and finally the North-West connection that runs through Guadalajara and Hermosillo to Mexicali. To transfer information between Guadalajara and Monterrey, for example, or between Monterrey and Mexicali, the Federal District must get involved, since there are no direct lines between those cities (COFETEL 1998b; 1999d; Iracheta Carroll 1999, 134; Red Tecnólogica Nacional 2000).
Centralizing infrastructure, Mexico City dominates the use of telecommunications. Though available data from the "Matrix Information and Directory Services" (MIDS 1999) on the distribution of hosts within Mexico are not very detailed - only orders of magnitude (for example: between 100,000 and 1,000,000 hosts) are reported -, they do show huge differences in scale among Mexican cities. In January 1999, Mexico City was the only Mexican city qualifying for the category "100,000 to 1,000,000 hosts", making it to one of the three telecommunications "hubs" in Latin America (the other two are São Paulo and Buenos Aires). Thus, Mexico City accounted for at least 50% of all hosts in Mexico, given that by that time there were slightly more than 200,000 hosts in the whole country. The centrality of Mexico City in the country's telecommunications network is shown as well by the fact that only one further city - surprisingly Reynosa - counts between 10,000 and 100,000 hosts, while other major cities such as Monterrey, Puebla, Chihuahua or Guadalajara have even less. In terms of Internet Service Providers (ISPs) and Points of Presence (PoPs)23, Mexico City's dominance is not that striking, although still remarkable. The Federal District houses 93 of the country's ISPs, while the federal states of Jalisco, Nuevo León and the State of Mexico have 45, each24. Likewise, nearly one third of the 395 Mexican PoPs are located in the Federal District (COFETEL 1998d and information provided by Oscar A. Robles). Finally, the geographical unevenness of the telecommunications system is revealed by the fact that 60% of all Mexican municipalities did not even have one single computer in 1995. Thus, in federal states like Oaxaca, Yucatán, San Luis Potosí, Chiapas or Guerrero between 93% and 71% of the municipalities lacked the very basic infrastructure to participate in the cross-border flows of information (Gutiérrez/Daltabuit 1999, 22).
Telecommunications are unevenly distributed not only at the national level. Rather, infrastructure and use are highly concentrated within the city, too. Basically, both the 2 megabits fiber optic line and the 128 and 64 kilobit lines are centralized in only five of the 16 "delegaciones" (districts) of the Federal District - Álvaro Obregón, Miguel Hidalgo, Benito Juárez, Cuauhtémoc and Coyoacán (Red Tecnólogica Nacional 2000). That comes with no surprise given that in these districts the main banks and companies, the research centers and universities, and many governmental institutions are located. 72% of the Federal Districts GDP in services originates in those five "delegaciones", 49% of the GDP in trade, and 21% in manufacturing (Iracheta Carroll 1999, cuadro 3). That shows that the integration in global data flows is highly concentrated both spatially and in terms of economic sectors.
Unfortunately, data on information flows which link Mexico City to other cities do not exist. However, all of the technological facilities alluded to implicitly inform about connectivity. Thus, evidence allows for two main hypothesis. First, Mexico is indeed becoming part of global telecommunications flows, as suggested by the sharp increase in international calls and use of Internet (which is, per definition, a cross-border telecommunications system). Second, all data available point to a new form of centrality, and that both at state and city level. Thus, if ones says that Mexico is integrated into global data flows it is meant that the Federal District, and even more precisely, a few areas of the Federal District, are connected to the cross-border telecommunications system. To trace links between these areas and cities elsewhere in the world in order to map global telecommunications flows remains a task for future research.
Connections between cities are also established by air travel. Though perhaps not as fundamental to the world economy as financial and commodity flows or telecommunications, air travel is significant because it reflects the ways in which cities are linked by the movement of people. Analyzing air travel between 22 world cities between 1985 and 1997, Smith et al. (in this volume; Smith/Timberlake 1995) found that Mexico City belongs to the third of four tiers of world cities. Both in 1985 and 1997, it ranked 16th in terms of the number of mutual relations with other cities in the network and in terms of the strength of the connections. Mexico City is more central than Miami, Montreal or Houston, but behind San Francisco, Chicago, Madrid or Zurich. However, the margin to the second tier cities is narrow25. Though its position in the network formed by air travel remained unchanged, Mexico City's power score (a term which Smith et al.[in this volume] use to measure inter-city relations by air travel) increased notably. While in 1985 it amounted to 0.16, by 1997 the power score of Mexico City had grown to 0.29. This means that Mexico City increased its share in the global airline traffic, which points to a growing integration in the global economy. The increase happened mainly in the second half of the 1980s - the time after Mexico joined the GATT (WTO). NAFTA seems to have had no impacts neither in terms of ranking nor in terms of power score of Mexico City. Nevertheless, in very recent years international air travel to the airport of Mexico City increased by nearly 10% annually, pointing to an again growing integration of the city into the global airline network (Secretaría de Turismo 1998)
Regarding the spatial patterns of Mexico City's integration in the international air travel network, it is remarkable to which extent U.S. cities dominate. In 1997, slightly more than two thirds of all international passengers came from or flew to U.S. cities (67.8 and 67.2%, respectively). About 12% of the passengers came from or flew to Europe, and about 11% connected to Latin American cities. How uneven the integration of Mexico City into the international air travel network is, can be illustrated by the fact that there was more air passenger travel between Mexico City and Los Angeles (amounting to 13% of all passengers) than between Mexico City and all European or Latin American cities. Next after Los Angeles come Houston, Dallas, Miami, New York and Chicago. Madrid ranks 7th, San José as top ranked Latin American city does not figure among the Top 10 (although connections to Latin America grew in recent years, while the dominance of the links to the U.S. cities decreased slightly). In terms of air cargo, the dominance of U.S. cities is not that impressive, although still strong. 55% of all freight originating in Mexico City went to U.S. cities, being again Los Angeles (17%) the most important destination, followed by Miami and New York. European cities amount together to a share of 22% (Paris is the most important one, followed by Frankfurt), while Latin American cities remain relatively unimportant as destinations of cargo from Mexico City (13%) - an amount that corresponds to only 75% of the cargo flown to Los Angeles. In terms of cargo flights to Mexico City, U.S. cities are far less outstanding - 42% of all freight came from there, compared to 36% of European and 16% of Latin American cities. Miami is the most important city of origin (13%), followed by Paris, Amsterdam and Los Angeles (Iracheta Carroll 1999, cuadros 4-7).
The finding that Mexico City is a rather central place in the network of cities formed by air travel corresponds to the above cited studies on advanced producer services and on telecommunications connections. However, airline data are somewhat problematic, because they include so different kinds of passengers as business men, tourists, or migrants. Though airline data provide good estimates of connections between cities established by the movement of people, they fail to grasp the qualitative difference of links. In the Mexican case a good part of the huge volume of flights to and from U.S. cities may in fact arise from migrants moving between the two countries and not from representatives of corporations or of national Chambers of Industry and Commerce. Indeed, the new airline Taesa basically came into existence to serve the migrant market, and Mexicana and AeroMexico have added flights in response. In this context it is interesting to note that Los Angeles is not only by far the most important city in terms of air passenger travel, but also houses more than a fourth of recent Mexican immigrants to the United States. Nearly half of all recent immigrants headed to five cities - Los Angeles, Chicago, New York, Houston and Phoenix (Durand et al. forthcoming [a], table 4) - the share of which in the air passenger travel amounts to more than a third. Although most of the "migrant flights" do not originate in the Federal District, but in cities of the west of the country (León, Guadalajara, Zacatecas, San Luis Potosí)(personal information by Douglas Massey), Mexico City became more important as a sending area for migrants to the U.S. in recent years (Durand et al. forthcoming [b]). Thus, both the amount and the spatial pattern of passenger air travel between U.S. cities and Mexico City might be influenced by migration.
If that were true, one would have to deal with a different quality of connections than in the case of, for example, air travel between London and New York. Yet, this does not mean that data on air travel of migrants are irrelevant, pointless to the question of connections between cities. In fact, the export of labor is one of the most important mechanisms of the integration of Mexico into the global economy, and Mexican emigrants have created firm links between the two countries. These connections might even surpass trade or foreign investment in economic significance (Secretaría de Relaciones Exteriores 1997, 31; Hinojosa Ojeda et al. 1998, 3). Thus, even if data on air travel do not refer to people who, in the words of Smith and Timberlake (1995, 296), are "greasing the wheels of production, finance or commerce through face-to-face contact" (emphasis added) but rather to Mexican men and women who grease the wheels of production, finance or commerce by really greasing machines, digging ditches for the cables for advanced telecommunications networks or cleaning bathrooms, air travel data support the argument that Mexico City is highly connected to U.S. cities.
Given the evidence of both strong linkages from Mexico City to other global cities and of specific urban transformations resulting from the city's new role in the world economy, it is obvious that "mega-city" is by no means a satisfactory term to grasp the particularity of Mexico City. There is much more to Mexico City than just a huge "Third World" urban agglomeration - without downplaying chronic problems (lack of satisfactory housing, pollution) or the more recent problem of crime, and without underestimating the significance of phenomena which have nothing to do with globalization, such as the demographic transition. What one observes is the making of a global city in some parts of the city - which enclose not the historical center that is dominated by public services and informal street vendors. The new "central business district" is spread to areas around Paseo de Reforma, Avenida Juárez, Santa Fe, Polanco, Insurgentes Sur, and Periférico Sur. Thus, the transformation of Mexico City towards a global city implies not only the emergence of new forms of centrality, but also the creation of new spaces of centrality within the city (Terrazas 2000).
Basically, the new spaces of centrality are located in five "delegaciones" of the Federal District: Miguel Hidalgo, Cuauhtémoc, Álvaro Obregón, Benito Juárez and Coyoacán. It is in these districts where nearly two thirds of the capital's GDP are produced (but three fourth of GDP in services), where the vast majority of companies that are headquartered in the Federal District can be found, where the conversion of the use of urban space towards services is most pronounced, where infrastructure for global telecommunication is centralized, where most of private investment is directed to, where many of the urban mega-projects (like shopping malls or modern office complexes) are (going to be) realized, and where prices for land and real estate are the highest. Interesting to note that the districts of Cuauhtémoc and Miguel Hidalgo are also the ones with the highest number of street vendors in Mexico City, and Benito Juárez is among the districts with the sharpest increase in informal trade. Thus, "center" and "periphery" in terms of economic activities and capital-labor relations are produced and reproduced on a very small spatial scale. The "new center" needs and involves the urban peripheries (in economic, social and spatial terms) since they are functionally linked to the core. An indigenous woman, living in a poor neighborhood like Valle de Chalco and working as a "muchacha" (domestic servant) in a banker's household in Las Lomas is attached to the "global" as well as a street vendor who sells branded articles. Both contribute in a specific niche to the reproduction of global capitalism26. Though rich studies on the links between the informal and the formal economy exist (see for example Benería/Roldán 1992), it remains a subject for further research whether the transformations Mexico City is undergoing create a section of society which is functionally irrelevant to the system - as Castells (1991, 213) put it in the context of whole states.
WHICH PLACE IN THE GLOBAL URBAN NETWORK?
Given the evidence that clearly shows that Mexico City is indeed a world city and as such well integrated into the cross-border urban network, two questions arise: First: Do the functions, which Mexico City carries out in and for the world economy, exceed the national territory and economy? Differently put, is Mexico City a regional center for Central America? Second: To which other world cities is Mexico City linked, and what is, in terms of power relations, the nature of these connections?
In Latin America, there are five cities identified by the "Globalization and World Cities Study Group and Network" as world cities: Mexico City, São Paulo, Buenos Aires, Caracas and Santiago. Since a majority of Latin American countries do not possess world cities, their access to the specialized knowledges of advanced producer services has to be cross-border. Could it be, in the case of Central American and (Spanish spoken) Caribbean countries, that Mexico City were the place where these services are bought by, for example, someone who wants to do business with Guatemala or Cuba? Evidence available suggest that that will not be the case. According to Taylor (2000), "where the region is the Americas, New York is the centre but where Latin America is a designated region then Miami is the centre". This argument is based on the fact that six out of eleven major producer services firms have their regional office in Miami, what makes that city to the clear regional center although not being a particularly important world city in its own right. Nijman (1996) reaches to the same conclusion. Although Miami does not play an important role neither on the national scene in the United States nor in trade relations with most of the world's regions, it is the dominant city regarding U.S. - Latin American connections. In the early 1990s, the city handled more than one third of all U.S. trade with Latin America, and the tendency was growing. For Central America and the Caribbean the share was even higher (47 and 43%, respectively). Yet, it is crucial to note that these data exclude Mexico. Regarding U.S. - Mexican trade, Miami virtually plays no role - it handles less than 1%. Thus, one can conclude that the reach of Mexico City as a global city does not exceed the national economy and territory, given that Miami seems to be the central node for the rest of Latin America. On the other hand, regarding the articulation of "Mexico" into the world economy, Miami is not involved. Consequently, in that case the main node is Mexico City.
From that further questions for future research can be derived: which cities have the most important links to the Federal District, and what is the nature of the links? Breaking down these rather general agenda, future investigations should focus on the following topics. First: Financial flows are, without any doubt, one of the most important mechanisms to tie Mexico and Mexico City to the global economy. In this regard, the city to which the Federal District is connected to is easy to identify. Financial markets in Mexico are intimately linked to - and dependent from - the New York stock exchange. Any "adjustment" of the Dow Jones would have serious effects on the Mexican financial system and the on the current account, on investments and on monetary policy, on interest rates and on the budget (El Financiero, 17.1.2000). Yet, the link to New York might not be the only one. Are there, for example, other stock markets which list Mexican companies? Second: The Federal District specializes, as shown above, on organizing, controlling and servicing foreign trade. Yet, we should know more about other cities involved in commodity chains that connect Mexico to the world market. Where else than in the Federal District is U.S. - Mexican trade serviced? Where to are commodities delivered? It is highly probable that the majority of these links are with U.S. cities, since 80% of Mexican's foreign trade is realized with the United States (CEPAL 2000, 105). Third: Although oil exports represent a shrinking part of all Mexican exports (about 6% in 1998 [BANCOMEXT 1999]), they still matter, particularly in regard of the federal budget. The headquarter of the national oil company Pemex is located in the Federal District, with connections to the oil producing states (like Campeche in the South-East of the country). Again, it is a subject for further research to identify other cities involved in the commodity chain - from trade partners to places where the oil prices are fixed. Fourth: Given the high dependence of Mexico on trade with and investment from the United States, the general performance of the U.S. economy is crucial for the destiny of the Mexican economy. For example, a slowing down of economic growth in the U.S. by one percentage point would result in a decrease of the GDP growth in Mexico by 0.7% (El Financiero, 17.1.2000). Alike, a sharp raising of interest rates in the U.S. would limit the possibilities for the Mexican state and Mexican companies to get external credits, and it would possibly make internal credits more expensive, too. All these factors would have impacts on financial flows to Mexico, on Mexican exports, and on domestic production. They are therefore likely to bring about an alteration in the role of Mexico City as a global city. Thus, it is important to know which cities are decisive for the fate of the U.S. economy, although there might be no direct links to the Federal District.
Regarding the nature of the integration of Mexico City into the cross-border urban system, links that connect cities can be distinguished by form and function, and also by power-related characteristics such as frequency, strength, importance or dominance/subdominance (Smith/Timberlake 1995, 290). Because links between cities are not even, power relations between cities - and nation-states - are crucial. Indeed, global city research stresses not only the function of and the connections among cities. There is also a strong notion of influence and hegemony. For Sassen (1991, 3) global cities are "highly concentrated command points in the organization of the world economy", and Knox (1995, 7) summarizes that they are "powerful centres of economic and cultural authority within the contemporary world-system". While global functions and power might be rather similar in cities like New York, London or Tokyo, in regard to "Third World" cities there is a marked inconsistency between function and power. Although Mexico City carries out important tasks in globalization processes, it is obvious that it can by no means compare to the power exercised in New York.
Although a study of power-related characteristics of cross-border links between Mexico City and other world cities lack, some general indications can be given. The Mexican stock market, for example, attracts only 0.005 % of the capital invested worldwide in stocks and bonds, and its turnover is only 1.6 % of that realized at Wall Street - a stock market to which the Mexican financial market is pegged (La Jornada, 15.4.1997, 24.8.1997; El Financiero, 14.4.1997). This quantitative dimension reflects, to a certain extent, a qualitative order - namely one of subordination. The Federal District as a financial center and Mexico - as an economy and as a state - are highly dependent on and vulnerable vis-a-vis the floating "global" capital. Both dependency and vulnerability are present in everyday life as shown by the financial crash in 1994-95, various devaluations, the contraction of wages, the sustained economic crisis, etc. (Dabat 1995; García 1995; Dussel Peters 1997). The uneven distribution of power in inter-city relations can be illustrated further by the fact that half of the biggest firms (in terms of sales and of exports) located in the Federal District are actually local headquarters of transnational corporations. In order to refine our knowledge of power-related characteristics of the cross-border links between Mexico City and other world cities, we should know to which extent decisions concerning the Mexican branch of these companies are met (or at least influenced) by the local offices. Alike, we should know which kind of services these corporations bring with them (and hence obtain them in other cities) and which ones are acquired locally. There is only one "global player" (companies among the world's "Top 500" firms listed by the journal Fortune) that has its "real" headquarter in Mexico City - the (still) state-owned oil-company Pemex. Among firms headquartered in the Federal District and dominated by private national capital, only a limited number operate on the world market (for example, Telmex, Carso, Bimbo, Cintra, Desc, Modelo)(Expansión, 21.7.1999; Chudnovsky et al. 1999, 174-179).
Consequently, Mexico City is not just another world city, neither. The fact that the terms "First" and "Third World" are not appropriate anymore to ascribe the geography and social hierarchy of the world - if they ever were - does not imply that "center" and "periphery" have become a pointless terminology. On the contrary; as socio-economic categories that shape the world system's hierarchy, "center" and "periphery" do still exist. Yet, they have undergone significant changes. In the era of globalization, "center" and "periphery" are produced and reproduced both on a global and a local scale. That makes it complicated to indicate a place for Mexico City in the global hierarchy. One the one hand, evidence presented clearly suggests that the city is no "center" in global terms. On the other hand, Mexico City does have core-areas (and people and activities), which form part of the cross-border network of cities. Thus, one might call Mexico City a "relay" global city - oriented to and dominated by one, probably tow powerful pole(s) and linked to other "relay cities" (compare Braudel 1986, 22-33). "Relay cities" like Mexico City (or São Paulo or Buenos Aires, etc.) are, as a group and as a category, indispensable for the whole system to exist. As shown for Mexico City, they do not only transmit global flows, but rather generate globalization on their own. However, as single, individual cities, "relay cities" might be nonessential - different to the "Alpha World Cities" identified by the "Globalization and World Cities Study Group and Network".
Five main conclusions can be drawn from this article. First, the specific way in which the city - and the country - are integrated into the international division of labor matters. For example, the character of the severe crisis which hit Mexico City in the 1980s can only be understood by taking into account the impacts resulting from the crisis of global capitalism. Thus, a purely urban or national approach is inadequate to grasp the recent economic, social and spatial transformations in Mexico City. Second, Mexico City is a crucial part of the cross-border network of cities. The Federal District is well integrated into global flows of capital, services, information and people, fulfilling global city functions and working more and more as a "hinge" between the "national" and the "global" economy. Consequently, one indeed observes the making of a global city. Third, this specific role of Mexico City in the global economy affects the urban economy, society and space. Though not all changes can be attributed to the making of a world city, recent transformations such as the boom of producer services are closely linked to globalization processes. Fourth, although evidence presented here supports the argument that Mexico City is (becoming) a global city, the challenge is still to prove it. Most, if not all, of the issues touched upon need to be investigated further in order to verify, modify or falsify the main hypothesis. Consequently, this article includes not only the evidence available, but also sketches out an agenda for future research. Fifth, and concluding, since there is much more to Mexico City than a huge urban agglomeration, one should reconsider the concept of "mega-cities". On the one hand, this term is highly problematic because of alluding so strong to population size which leads, vice versa, to a neglect of qualitative new urban dynamics. On the other hand, global city research provides, as shown in this article, very useful tools which could help us to deepen our knowledge of the major cities of the "Third World".
Aguilar, Adrián Guillermo. 1996. Reestructuración económica y costo social en la Ciudad de México. Una metrópoli periférica en la escena global. Ponencia presentada en el Seminario "Economía y Urbanización: Problemas y Retos del Nuevo Siglo", organizado por el Instituto de Investigaciones Económicas, UNAM, en la Unidad de Seminarios "Dr. Ignacio Chávez", 20-22 de mayo 1996.
Aguilar, Adrián Guillermo, Boris Graizbord, Álvaro Sánchez Crispín. 1996. Las ciudades intermedias y el desarollo regional en México. Consejo Nacional para la Cultura y las Artes: México D.F.
Altvater, Elmar, Birgit Mahnkopf. 1996. Grenzen der Globalisierung. Ökonomie, Ökologie und Politik in der Weltgesellschaft. Westfälisches Dampfboot: Münster.
Amin, Samir/Giovanni Arrighi, André Gunder Frank/Immanuel Wallerstein. 1982. Dynamics of Global Crisis. Monthly Review Press.
Armstrong, Warwick/T.G. McGee: 1985 Theatres of Accumulation: Studies in Asian and Latin American Urbansisation. Methuen, London.
Arrighi, Giovanni: 1994 The Long Twentieth Century. Money, Power, and the Origins of Our Times. Verso. London.
Bancomext: 1999 Comercio Exterior de México, http://www.bancomext.gob.mx/espa/com_ext/index_est.html
Bancomext 2000: Indice de Precios; http://www.bancomext.gob.mx/espa/com_ext/index_est.html
Bataillon, Claude. 1992. Servicios y empleo en la economía de la ZMCM. In: Consejo Nacional de Población: La Zona Metropolitana de la Ciudad de México. Problemática actual y perspectivas demográficas y urbanas. CONAPO: México D.F., 79-83.
Beaverstock, J.V., R.G. Smith and P.J. Taylor: 1999a Geographies of Globalization: US Law Firms in World Cities. In: GaWC Research Bulletin 4.
Beaverstock, J. V./R. G. Smith/P. J. Taylor: 1999b A roster of world cities. In: Cities, 1999, 16 (6), 445-458
Beaverstock, J. V./R. G. Smith/P. J. Taylor: 1999c The global capacity of a world city: a relational study of London. In: GaWC Research Bulletin 7, edited and posted on the Web on 28th July 1999.
Beavon, Keith: 1998 "Johannesburg": Coming to grips with globalization from an abnormal base. In: Lo, Fu-Chen/Yue-Man Yeung (eds): Globalization and the world of large cities. UNU Press. Tokyo, 352-387.
Bendesky, León: 2000 La industria maquiladora. In: La Jornada, 25.1.2000, 17.
Benería, Lourdes, Martha I. Roldán. 1992. Las encrucijadas de clase y género. Trabajo a domicilio, subcontratación y dinámica de la unidad doméstica en la ciudad de México. El Colegio de México, Fondo de Cultura Económica: México D.F.
Boltvinik, Julio. 1995. La Satisfacción de las Necesidades Esenciales en México en los Setenta y Ochenta. In: Garza, Luis Alberto de la, Enrique Nieto (eds): Distribución del Ingreso y Politicas Sociales. Tomo I. Semenario Nacional Sobre Alternativas Para la Economía Mexicana. Juan Pablos Editor: México D.F., 17-77.
Braudel, Fernand: 1986 Sozialgeschichte des 15.-18. Jahrhunderts. Aufbruch zur Weltwirtschaft. Kindler. München.
Camposortega Cruz, Sergio. 1992. Evolución y tendencias demográficas de la Zona Metropolitana de la Ciudad de México. In: Consejo Nacional de Población: La Zona Metropolitana de la Ciudad de México. Problemática actual y perspectivas demográficas y urbanas. CONAPO: México D.F., 3-15.
Cárdenas Solórzano, Cuauhtémoc: 1999, Segundo informe de gobierno. Anexo Estadístico. Gobierno del Distrito Federal. Cd. de México.
Castells, Manuel: 1991 Die zweigeteilte Stadt - Arm und Reich in den Städten Lateinamerikas, der USA und Europas. In: Tilo Schabert (ed): Die Welt der Stadt. München/Zürich 1991, S 199-216.
Chase-Dunn, Christopher K. 1985: The System of World Cities, 800 A.D.-1975. In: Timberlake, Michael (ed.): Urbanization in the World-Economy. Academic Press: Orlando. 269-292.
Chudnovsky, Daniel/Bernardo Kosacoff/Andrés López: 1999 Las multinacionales latinoamericanas: sus estrategias en un mundo globalizado. Fondo de Cultura Económica. México D.F.
Comisión Económica para América Latina y el Caribe (CEPAL): 2000 La inversión extranjera en América Latina y el Caribe. CEPAL. Santiago de Chile.
Comisión Federal de Telecomunicaciones (COFETEL) 1998a Cronología de la telefonía en México. http://www.cft.gob.mx/html/la_era/info_tel2/hist8.html
Comisión Federal de Telecomunicaciones (COFETEL) 1998b Estadísticas de Interés sobre Telecomunicaciones: RTN: Backbone + Nodos regionales. http://www.cft.gob.mx/html/5_est/graficas/img038.gif
Comisión Federal de Telecomunicaciones (COFETEL) 1998c Estadísticas de Interés sobre Telecomunicaciones: Conferencias de larga distancia. http://www.cft.gob.mx/html/5_est/graficas/Graf4_pag6.html
Comisión Federal de Telecomunicaciones (COFETEL) 1998d Estadísticas de Interés sobre Telecomunicaciones: Estados que cuentan con más de 30 proveedores ISP's. http://www.cft.gob.mx/html/5_est/Graf_internet/proveedores_isps.htm
Comisión Federal de Telecomunicaciones (COFETEL) 1999a Estadísticas de Interés sobre Telecomunicaciones: Líneas Telefonía en Servicio y Densidad Telefónica. http://www.cft.gob.mx/html/5_est/celulares/usumincel.html
Comisión Federal de Telecomunicaciones (COFETEL) 1999b Estadísticas de Interés sobre Telecomunicaciones: Porcentaje de Digitalización de la Planta Telefónica. http://www.cft.gob.mx/html/5_est/graficas/Graf3_pag5.html
Comisión Federal de Telecomunicaciones (COFETEL) 1999c Estadísticas de Interés sobre Telecomunicaciones: Usarios estimados de Internet en México http://www.cft.gob.mx/html/5_est/Graf_internet/estiminternet_01.html
Comisión Federal de Telecomunicaciones (COFETEL) 1999d Estadísticas de Interés sobre Telecomunicaciones: Lineas Residenciales, Comerciales y Totales. http://www.cft.gob.mx/html/5_est/graficas/lcomres.html
Connolly, Priscilla. 1993. La reestructuración económica y la ciudad de México. In: Coulomb Bosc, René, Emilio Duhau (eds): Dinamica urbana y procesos socio-politicos. Lecturas de actualización sobre la Ciudad de México. UAM-Azcapotzalco: México D.F., 45-70.
Connolly, Priscilla. 1997. Cuál megalópolis? Ponencia presentada en el Congreso Internacional Ciudad de México, Sobre Política y Estudios Metropolitanos, organizado por el Consejo Mexicano de Ciencia Sociales, A.C., 10 al 14 Marzo.
Consejo Nacional de Población (CONAPO): 1999a Proyecciones de la población de las entidades federativas y sus municipios, 1995-2030. CONAPO. México D.F.
Consejo Nacional de Población (CONAPO): 1999b La situación demográfica de México. CONAPO. México D.F.
Consulado General de México en Nueva York/ SHCP-BANXICO-INEGI: 2000 Comercio Exterior de México http://www.quicklink.com/mexico/
Corona Cuapio, Reina/Rodolfo Luque González: 1992 El perfil de la migración de la Zona Metropolitana de la Ciudad de México. In: Consejo Nacional de Población: La Zona Metropolitana de la Ciudad de México. Problemática actual y perspectivas demográficas y urbanas. CONAPO, México D.F., 21-31.
Corona, Rodolfo/Rodolfo Tuirán: 1994 Profundas transformaciones regionales. In: Demos. Carta demográfica sobre México, No. 7, 21-22.
Dabat, Alejandro. 1995. La crisis mexicana y el nuevo entorno internacional. In: Comercio Exterior, núm 11, 866-874.
Davis, Diane E.. 1993. Crisis fiscal urbana y los cambios políticos en la ciudad de México: desde los orígenes globales a los efectos locales. In: Estudios Demográficos y Urbanos 22, vol. 8, núm. 1, 67-102.
de la Peña, Guillermo, J.M. Duran, Augustín Escobar, J. García de Alba (eds): 1990 Crisis, conflicto y sobrevivencia. Estudios sobre la sociedad urbana en México. Guadalajara. Editorial Universidad de Gualdalajara.
de Mateo, Fernando. 1998. NAFTA; Foreign Direct Investment and Economic Integration: The Case of Mexico. Paper prepared for the Seminar on Migration, Free Trade and Regional Integration in North America, organised by the OECD and the Mexican Authorities with the Support of Canada and the United States. Mexico City, 15-16 January 1998.
Delgado Selley, Orlando: 2000 Resultados macroeconómicso y nivel de vida. In: La Jornada, 28.1.2000
Drakakis-Smith, David (ed): 1986 Urbanisation in the Developing World. Routledge. London.
Durand, Jorge/ Douglas S. Massey/ Fernando Charvet: forthcoming (a). The Changing Geography of Mexican Immigration to the United States: 1910-1996. In: Social Science Quarterly
Durand, Jorge/Douglas S. Massey/René M. Zenteno: forthcoming (b). Mexican Immigration to the United States:Continuities and Changes. In Latin American Research Review
Dussel Peters, Enrique. 1997. La Economía de la Polarización. Teoría y Evolución del Cambio estructural de las Manufacturas Mexicanas (1988-1996). UNAM: México D.F.
Dussel Peters, Enrique: 1999 La inversión extranjera en México. Informe 1999. Documento elaberado para la Comisión Económica para América Latina y el Caribe (CEPAL).
Enloe, Cynthia: 1990 Bananas, Beaches & Bases. Making Feminist Sense of International Politics. Berkeley and Los Angeles: The University of California Press.
Feldbauer, Peter/Erich Pilz/Dieter Rünzler/Irene Stacher: 1993 Megastädte. Zur Rolle von Metropolen in der Weltgesellschaft. Böhlau. Wien.
Fernández, Jeffry S.: 1995 Development of WWW services in Mexico, toward a National Information Infrastructure. http://www.isoc.org/HMP/PAPER/035/html/paper.html
Frank, André Gunder. 1990. Politische Ironien der Weltwirtschaft. In: Frank, André Gunder/Marta Fuentes-Frank: Widerstand im Weltsystem. Kapitalistische Akkumulation, Staatliche Politik, Soziale Bewegung. Wien. Promedia, Wien, 13-45.
Friedmann, John: 1986 The World City Hypothesis. In: Development and Change, 17, 69-83.
Friedmann, John. 1995. Where we stand: a decade of world city research. In: Knox, PaulL., Peter J. Taylor (eds): World cities in a world system. Cambridge: Cambridge University Press, 21-47.
García, Samuel: 1995 Corrientes de capital y lecciones de la crisis financiera mexicana. In: Comercio Exterior, num 12, 932-935.
Garza, Gustavo: 1985 El Proceso de Industrialización en la Ciudad de México (1821-1970). El Colegio de México: México D.F.
Garza, Gustavo, Salvador Rivera: 1994 Dinamica Macroeconomica de las Ciudades en México. Instituto Nacional de Estadísticas, Geografía y Informatica: Aguascalientes.
Gilbert, Alan. 1992. Urban Development in a World System. In: Gilbert, Alan/Josef Gugler (eds.): Cities, Poverty and Development: Urbanization in the Third World. Oxford University Press: Oxford. 14-32.
Gómez de León Cruces, José/Virgilio Partida Bush: 1996 La Ciudad de México: tendencia demográfica y escenarios para el siglo XXI. In: Federalismo y Desarollo, 9, 56, 12-17.
Gugler, Josef: 1999 Introduction. Paper prepared for a National Academies Workshop on World Cities in Poor Countries. Washington DC, October 27-28th, 1999.
Gutierrez, Fernando, Enrique Daltabuit: 1999 Mexican cities in cyberspace. In: Cities, 1999, Vol 16, Iss 1, 19-31.
Hiernaux Nicolás, Daniel: 1994 Hacia la ciudad neoliberal? Algunas hipótesis sobre el futuro de la ciudad de México. In: Hiernaux Nicolás, Daniel, Francois Tomas: Cambios económicos y perifería de las grandes ciudades. El caso de la ciudad de México. UAM-IFAL: México D.F, 22-45.
Hiernaux Nicolás, Daniel. 1995. Reestructuración económica y cambios territoriales en México. Un balance 1982-1995. In: Estudios Regionales, núm. 43, 151-176.
Hinojosa Ojeda, Raul/Robert McCleery/Fernando de Paolis: 1998 Economic effects on NAFTA: Employment and Migration Modelling Results. Paper prepared for the Seminar on Migration, Free Trade and Regional Integration in North America, organised by the OECD and the Mexican Authorities with the Support of Canada and the United States. Mexico City, 15-16 January 1998.
Instituto Nacional de Estadísticas, Geografía y Informatica: 1994 Estadisticas Historicas de México. Tomo I. INEGI: Aguascalientes.
Instituto Nacional de Estadísticas, Geografía y Informatica: 1995 Migración Reciente en México, 1985-1990. INEGI, Aguascalientes.
Instituto Nacional de Estadísticas, Geografía y Informatica. 1996. Estados Unidos Méxicanos. Conteo de Población y Vivienda 1995. Resultados Definitivos. Tabulados Basicos. INEGI: Aguascalientes.
Instituto Nacional de Estadísticas, Geografía y Informatica: 1998 Anuario estadístico del Estado de México. INEGI, Aguascalientes.
Instituto Nacional de Estadísticas, Geografía y Informatica: 1999 Estadisticas Economicas. Indicadores de empleo y desempleo. INEGI. Aguascalientes.
Instituto Nacional de Estadísticas, Geografía y Informatica. 2000a. Banco de Información Económica, Producto interno bruto trimestral. http://www.inegi.gob.mx/
Instituto Nacional de Estadísticas, Geografía y Informatica. 2000b. Banco de Información Económica, Sector Externo, Balanza de Pagos de México, Cuenta Corriente, http://www.inegi.gob.mx/
Instituto Nacional de Estadísticas, Geografía y Informatica. Various years (a). Sistema de cuentas nacionales en México. Producto Interno Bruto por entidad federativa 1993. INEGI, Aguascalientes.
Instituto Nacional de Estadísticas, Geografía y Informatica. Various years (b). Encuesta Nacional de Empleo Urbano. INEGI: Aguascalientes.
Instituto Nacional de Estadísticas, Geografía y Informatica. Various years (c). Encuesta de Ingresos y Gastos de los Hogares del Area Metropolitana de la Ciudad de México. INEGI, Aguascalientes.
Internet Software Consortium (ISC) 1999. Domain Survey. Distribution by Top-Level Domain Name by Host Count. http://www.isc.org/ds/WWW-9907/dist-bynum.html
Iracheta Carroll, Jimena del Carmen: 1999 Las grandes ciudades en el contexto de la globalización: El caso de la Zona Metropolitana del Valle de México. Tesis para sustentar el título de Licenciado en planeación territorial. Universidad Autónoma del Estado de México. Toluca.
Knox, Paul L.. 1995. World cities in a world system. In: Knox, PaulL., Peter J. Taylor (eds): World cities in a world system. Cambridge: Cambridge University Press, 3-20.
Knox, Paul L./Peter J. Taylor (eds): 1995. World cities in a world system. Cambridge University Press. Cambridge.
Livas Elizondo, Raúl A.: 1994 Desarollo regional y apertura comercial. In: Examen de la Situación Económica de México, 2, núm. 819, 85-91.
Lo, Fu-Chen/Yue-Man Yeung (eds): 1998 Globalization and the world of large cities. UNU Press. Tokyo.
Lustig, Nora: 1994 México. Hacia la reconstrucción de una economía. El Colegio de México: México D.F.
Matrix Information and Directory Services (MIDS) 1999 The Internet in Mexico, January 1999, http://www.mids.org/mmq/602/mid/intbmx.html
Molina Ludy, Virginia, Kim Sánchez Saldaña. 1997. El Fin de la Ilusión. Lo Movilidad Social Entre Familias de la Ciudad de México. Ponencia presentada en el Congreso Internacional Ciudad de México, Sobre Política y Estudios Metropolitanos, organizado por el Consejo Mexicano de Ciencia Sociales, A.C., 10 al 14 Marzo.
Nijman, J.: 1996 Breaking the rules. Miami in the urban hierarchy. Urban Geography, 17, 1, 5-22.
OECD. 1995. OECD Economic Surveys 1994-1995. Mexico. OECD. Paris.
OECD 1999 Economic Surveys. Mexico. OECD. Paris.
Parnreiter, Christof. 1998a. Migration in Megastädte der Dritten Welt. Von der importsubstituierenden Industrialisierung zur Globalisierung. Erfahrungen aus Mexiko. Dissertation zur Erlangung des Doktorgrades an der Geisteswissenschaftlichen Fakultät der Universität Wien. Wien.
Parnreiter, Christof. 1998b. Globalisierung und Megastädte der Peripherie. In: Journal für Entwicklungspolitik 4, 325-336.
Parnreiter, Christof: xxxx Free Trade and Changing Patterns of Cityward Migration: The Case of Mexico. In: xxxx see UNESCO Publication
Partida Bush, Virgilio: 1995 Migración Interna. Tomo II. Instituto Nacional de Estadísticas, Geografía y Informatica (INEGI), Aguascalientes.
Porras Macías, Agustín: 1997 El Distrito Federal en la dinámica demográfica megalopolitana en el cambio de siglo. In: Eibenschutz Hartman, Roberto (ed): Bases para la planeación del desarollo urbano en la Ciudad de México. Tomo I: Economía y sociedad en la metropoli. Universidad Autónoma Metropolitana - Xochimilco México D.F., 37-73.
Pradilla Cobos, Emilio. 1993. Territorios en crisis. México 1970-1992. Red Nacional de Investigación Urbana y Universidad Autónoma Metropolitana: México D.F.
Pradilla Cobos, Emilio. 1997. La Megalópolis Neoliberal: Gigantismo, Fragmentación y Exclusión. Ponencia presentada en el Congreso Internacional Ciudad de México, Sobre Política y Estudios Metropolitanos, organizado por el Consejo Mexicano de Ciencia Sociales, A.C., 10 al 14 Marzo.
Red Mexicana de Acción Frente al Libre Comercio. 1997. Espejismo y Realidad: El TLCAN Tres Años Después. Análisis y Propuesta Desde la Sociedad Civil. RMALC: México D.F.
Red Tecnólogica Nacional: 2000 Nodos. rtn.net.mx/infotec/RTN.html
Rivera, Salvador. 1997. Cambios en el desarollo urbano: ¿Es la globalización una era de desconcentración urbana? In: Demos. Carta demográfica de México, Núm. 10, 27-29.
Sassen, Saskia. 1991. The Global City. New York, London, Tokyo. Princeton University Press. Princeton.
Sassen, Saskia. 1996. Cities in the Global Economy. Paper presented at the UNU-Panel on Globalization and the Urban Future, June 7, Habitat II.
Secretaría de Relaciones Exteriores/Commission on Immigration Reform: 1997 Estudio Binacional México-Estados Unidos sobre Migración/Binational Study on Migration Between Mexico and the United States. México D.F. y Washington.
Secretaría de Comercio y Fomento Industrial (SECOFI). Dirección General de Inversión Extranjera. México D.F.
Secretaría de Turismo: 1998 Estadísticas básicas de la actividad turística. México D.F.
Short, J.R./Y. Kim/M. Kuus/H. Wells: 1996 The Dirty Little Secret of World Cities Research: Data Problems in Comparative Analysis. In: International Journal of Urban and Regional Research, vol. 20, number 20, 697-717.
Smith, David A./Michael Timberlake: 1995 Conceptualising and Mapping the Structure of the World System's City System. In: Urban Studies, vol. 32, no. 2, 287-302.
Taylor, Peter J.: 1997 Hierarchical tendencies amongst world cities: a global research proposal. In: Cities, 14, 323-332.
Taylor, Peter J.: 1999a So-called "World Cities": The Evidential Structure within a Literature. In: Environment and Planning A, 1999, 31 (11), 1901-1904.
Taylor, Peter J.: 1999b Worlds of Large Cities: Pondering Castells' Space of Flows. Globalization and World Cities Study Group and Network, Reserch Bulletin 14, October 1999
Taylor, Peter J.: 2000 World Cities and Territorial States under Conditions of Contemporary Globalization. In: Political Geography, 19, 1, 5-32.
Taylor, P.J./D.R.F. Walker: 1999 World Cities: A First Multivariate Analysis of their Service Complexes. In: GaWC Research Bulletin 13, edited and posted on the Web on 25th October 1999.
Terrazas, Oscar: 2000 La red de la centralidad metropolitana en la globalización. Borrador de trabajao. Mèxico D.F.
Timberlake, Michael (ed.): 1985 Urbanization in the World-Economy. Academic Press, Orlando.
Tolosa, Hamilton C.: 1998 Rio de Janeiro as a world city. In: Lo, Fu-Chen/Yue-Man Yeung (eds): Globalization and the world of large cities. UNU Press. Tokyo, 203-227.
Walton, John. 1976. Political economy of world urban systems, directions for comparative research. In: Walton, J./L. Masotti (eds.): The city in comparative perspective, cross-national research and new directions in theory. John Wiley &Son: New York.
Ward, Peter M.: 1998 Mexico City. Revised Second Edition. Wiley. Chichester.
Ziccardi, Alicia: 1998 Gobernabilidad y participación ciudadana en la Ciudad de México. Instituto de Investigaciones Sociales, UNAM. México D.F.
El Financiero, various numbers
*. Institute for Urban and Regional Research, Austrian Academy of Sciences, Vienna.
1. The author thanks the "Fonds zur Förderung der Wissenschaftlichen Forschung" (Vienna) for granting him an "Erwin Schrödinger Auslandsstipendium" to the University of Chicago. Additionally, he thanks both the "Globalization Project" and the "Transnationalism Project" of the University of Chicago for their support. For support I thank Enrique Dussel Peters, Peter Feldbauer, Adolfo Gilly, Karl Husa and Saskia Sassen. Comments are very welcome (Christof.Parnreiter@univie.ac.at).
2. "Mexico City" refers to the whole Metropolitan Area, or the "Zona Metropolitana de la Ciudad de México" (ZMCM), which in 1995 had, according to the "National Institute for Statistics, Geography and Computer Sciences" (Instituto Nacional de Estadística, Geografia e Informatica, INEGI), 16.7 million inhabitants. The ZMCM is formed by the "Federal District", which is the country's capital, and those communities ("municipios") of the surrounding State of Mexico (30 in 1995), which have grown together with the Federal District. In 1995, the Federal District covered less than one third of the city's surface and housed slightly more than 50% of its population (INEGI 1996; Iracheta Carroll 1999, 73; Camposortega Cruz 1992, 4). The "National Population Council" (Consejo Nacional de Población, CONAPO) applies a slightly different definition. According to CONAPO, Mexico City had 18.6 million inhabitants in 1996 and contained the Federal District and 58 "municipios" of the State of Mexico and one of the state of Hidalgo (CONAPO 1999a).
3. As early as in 1976 Walton called for a change in the perspective of urban research, arguing that it is largely economic forces operating cross-nationally that shape "Third World" urbanization. For early studies applying a global approach, see Timberlake (1985), Armstrong/McGee (1985) and Drakakis-Smith (1986).
4. Maquiladoras are in-bond-industries geared to foreign markets (mainly the U.S.) which are mainly located in the Northern border region and which process imported semi finished products to re-export them.
5. GDP growth in the Federal District was above even of the dynamic north-eastern region of the country (formed by the states of Coahuila, Nuevo León and Tamaulipas), but below of growth in states like Aguascalientes, Baja California, Chihuahua, Querétaro or Quintana Roo.
6. Of course, one always faces the problem of documentation (who to count and how?), as the frequent debate on census results show. However, the results of both the 1990 and the 1995 census are widely acknowledged among Mexican researchers. Yet, migration data presented by the national statistic institute INEGI suffer from another weakness. They include intra-urban mobility within Mexico City, that migration between the Federal District and one of the communities belonging to the State of Mexico, but physically tied to the Federal District (see Partida Bush 1995, 19). Once these statistical distortions are corrected, the number of real net-emigration from the ZMCM is considerably lower than the ones quoted by INEGI (for recalculation see Parnreiter 1998a, 118-130).
7. This was mainly due to an over-supply of the dollar. In the words of Arrighi (1994, 323): "First World bankers (were) begging Third World States to borrow their overabundant capital".
8. São Paulo, for example, experienced a crisis very similar to that of Mexico City (Tolosa 1998, 214-216).
9. In a similar vein, in South Africa and Brazil it is questioned whether the shift to export production will weaken or strengthen Johannesburg and São Paulo, respectively. Evidence so far points to a reinforcement of both Johannesburg's and São Paulo's position (Beavon 1998, 363f; Tolosa 1998, 217-219).
10. There seem to be no direct links between Mexico City and the maquiladora industry. Between 1993 and 1999, 77% of the production value of this industry corresponded to imports, and only 23% was value added in Mexico. Of this rather small segment, 51% correspond to wages (in the vast majority blue collar workers), 13% to profits, 8% to packing and row materials, and 27% to other factors (such as properties, energy, transportation etc.)(Bendesky 2000). The low value added in Mexico and its structure show that low skilled labor is in fact the only important Mexican contribution to the maquiladora industry. That allows for the hypothesis that specialized services (e.g. financial or legal services) that are be needed are imported - presumably from the United States. However, it might be that there are "indirect" links between the maquiladora industry and the Federal District. It might, for political reasons, be necessary for companies running maquila-plants to have representatives in Mexico City - for example, to do lobbying.
11. In terms of sales, GM comes third (behind Pemex and Teléfonos de México), Daimler-Chrysler is fourth, Volkswagen sixth and Ford eighth (1998). In terms of exports, GM is second (again behind Pemex), Daimler-Chrysler third, Volkswagen fourth and Ford sixth (data for Nissan are not avaiable). In terms of direct foreign investment, the most important investment projects in 1998-99 were made by Chrysler, Ford, Volkswagen and Nissan (Expansión, 21. Julio 1999; Dussel Peters 1999, cuadro 11).
12. New investment plus investment in maquiladoras.
13. Employment is registered by the National Survey on Urban Employment (Encuesta Nacional de Empleo Urbano), which covers 43 cities (INEGI, various years).
14. Data are for 1993; the poverty line is defined by an earning of three minimum wages.
15. Data on the whole ZMCM are not available.
16. The migration balance with the most important sending state, the State of Mexico, is negative at a very high level (-468,069). However, due to the complexity of the spatial situation (the State of Mexico forms part of the ZMCM) this figure is misleading. Yet, excluding intra-urban mobility, the migration balance of the Distrito Federal with the State of Mexico is still negative (-9,695).
17. It is interesting to note that Mexico City is not classified as a "Third World" but as "outer city" to the United States. Furthermore, it should be mentioned that the "Globalization and World Cities Study Group and Network" does not distinguish between Mexico City and the Federal District. It is likely that the data cited refer to the Federal District rather than to the whole urban agglomeration.
18. "To qualify as "major" global service center in, for example, banking, 5 to 7 of the 10 firms analyzed have to have significance presence in a city; among the 6 accounting firms surveyed, 3 to 4 firms have to have significance presence in a city, etc. A higher presence would result in being a "prime" global service center, a lower in a "minor" global service center.
19. Accountancy presents a particular case because it is the most concentrated of all producer services resulting in very large firms having the most widespread pattern of offices of all services. Thus, the five firms identified by the "Globalization and World Cities Study Group and Network" have branches in all cities, which translates into a 100% likelihood for all cities to have direct organizational links to London.
20. The ranking is based on technical definitions of the pricing of services, the choice of physical infrastructure connections available and the availability of the most advanced connections.
21. Not counting Hong Kong as a "Third World" City.
22. Defined as computers permanently connected to the Internet.
23. A PoP is the place where an ISP connects to its customers.
24. Since Mexico City stretches to the State of Mexico, it is highly probable that at least some of the ISPs in the State of Mexico are in fact located in the ZMCM.
25. Unfortunately, São Paulo is not included in the data on air travel between 1985 and 1997. In an earlier article Smith and Timberlake (1995) ranked it 22nd - markedly behind Mexico City.
26. There is another relation between the peripheries and the global that should be mentioned. As a consequence of the debt crisis and the neoliberal restructuring, ever more people are forced into migrations to compensate for the decrease in wages or the loss of income through agricultural activities. Thus, the indigenous woman who decided to emigrate in order to make money as a domestic servant "is designing her own international debt politics" (Enloe 1990, 185) by cleaning the bankers bathrooms.
Table 1: Location Patterns of the Main Firms Registered in Mexico, 1998 (in Percentage)
Sources: Own calculations, based on Expansión, 1999: Las empresas más importantes de México. Expansión S.A.
a) ZMCM (Zona Metropolitana de la Ciudad de México) defined as, according to the INEGI, the Federal District plus 30 "municipios" of the "Edo. de México" (see footnote 3).
Note: The three state owned enterprises, which are listed among the 500 most important Mexican firms, are not included into table 1. Petróleos Mexicanos (Pemex), the biggest firm registered in Mexico, has its headquarter in the Federal District, while the other two, which are of minor importance, are located in Baja California Sur. The table does not include data on maquiladoras.
Table 2: Distribution of Foreign Direct Investment in Mexico, 1989-1998
Source: 1989-1993: SECOFI. Dirección General de Inversión Extranjera; 1994-1998: Own calculations, based on Dussel Peters 1999, cuadros 8, 9. Note: Data on FDI between 1994 and 1998 include new investment plus investment in maquiladoras. Due to a change in the methodology to measure FDI, the absolute numbers before and after 1994 are not comparable. However, since the purpose of this article is to show the regional distribution of FDI, a comparison can be made.
Table 3: Employment in Real Estate, Financial and Professional Services, 1987-1997
Source: INEGI, various years (b).
Edited and posted on the web on 5th January 2000; last update 24th February 2000