The Central American region (taken here to encompass the five republics of Costa Rica, El Salvador, Guatemala, Honduras and Nicaragua) has been intensely integrated into the world economy since the nineteenth century. At that time it became firmly entrenched within the expanding international trading system as a provider of agricultural commodities. For most of the twentieth century the Central American republics can be characterised as archetypal agro-export societies whose economic fortunes were tied to the performance of the export sector. Thus openness to the world economy is certainly nothing new to this region: typically these five republics have demonstrated a higher reliance on international trade than other countries of equivalent size (Brown 1994,161-167).1
This situation bestowed an enormous amount of economic power on those elites who controlled the capital accumulation derived from the export trade. A particular social structure emerged to sustain the agro-export model that was both highly inegalitarian and harshly repressive (with the partial exception of Costa Rica).2 From the late 1960s, however, the interaction of a series of global and regional processes began to unravel this structure and set in motion a series of far-reaching transformations. These included the growing revolutionary challenge mounted to the traditional order which erupted in the late 1970s, changing US geopolitical interests in the region, the processes of global economic integration, and the eventual consolidation of a new US-sponsored model of Neoliberal adjustment in the Americas. This has been interpreted by William Robinson (1997a) as the supplanting of the traditional national agro-export model of Central American development by a new transnational one. New non-traditional products (Willmore 1997), tourism (Rodas-Martini nd, 35-38), and remittances from emigrant labour (Kaimowitz 1990; Orozco et al 1997) have been 'the most dynamic economic sectors linking the region to global circuits of production and distribution' (Robinson, 1997a: 31) in this new era.
Here we focus upon one particular aspect of this new transnational model through an examination of those linkages involving financial and other business services. From this respect, the Neoliberal policies of the past fifteen years that have liberalised commercial and financial markets and lifted restrictions on foreign investment across the five republics have been the specifically relevant reforms (Evans, 1998).3
In this paper we attempt to relate these important regional changes to wider spatial developments in the world economy. Economic globalization is producing a new geographical restructuring in which global spaces of flows are becoming dominant (Castells, 1996). In particular, we are concerned with the way in which the world economy is being articulated through a world city network of business service centres (Sassen, 1991, 1994). Readily identifiable by their office tower blocks, world cities house advanced producer service firms who operate through networks of offices to link cities together (Taylor, 2001). Today there are global-level services provided in a range of professional (e.g. inter-jurisdictional law), creative (e.g. world-wide advertising) and financial activities (e.g. global asset management) where firms strive to offer a 'seamless service' under their own brand name. For every service firm with global clients, this necessitates a network of offices in major cities across the world. There are numerous different rosters of these world cities (Beaverstock et al., 1999) but one feature they all have in common is that no Central American cities appear on any list. The leading cities in each of the five republics have neither the urban market nor regional hinterland to qualify for world city status. It is this circumstance of a region without a world city that is at the core of this study.4
Central America might be geographically beyond world cities but this does not mean that it is not serviced through world cities. Linkage to the outside world is, by definition, central to the region's transnational model of development and business services, and therefore world cities, are integral to this. Hence the objective of this paper can be cast in two different forms. First, from a regional perspective, we investigate specific ways in which Central America is integrated into the world economy through business services. Second, from a global perspective, we investigate specific ways in which a region beyond the world city network obtains its global business services. In other words this research can contribute to two different sets of literature. We think that both literatures have an empirical deficit when it comes to describing the specifics of the linkages that create the new spaces of flows in economic globalization. Thus what we provide is a preliminary exploration of a particular regional global integration into the world city network.
There are no official data that show the economic links between Central America and the world city network. Data have to be derived using information that is available from advanced producer service firms which have business links within and without Central America. There are two sources of such data used in this study.
First, we use data collected for a global analysis of world cities and business services and which is described in detail in Taylor et al. (2001). These original data consists of a matrix of the 'service values' of 100 global service firms5 across 316 cities worldwide. The service values are estimates of the importance of a city within a firm's global office strategy taking into account features such size and function of offices. Because of the multifarious nature of the data, drawn largely from the web sites of firms, service values are based upon a simple coding ranging from 0 (no presence) to 5 (headquarters). For this study we extract the data pertaining to Central America (and also Latin America more generally, see below), in particular focussing on the five Central American cities in the data: Guatemala City, Managua, San Jose, San Salvador and Tegucigalpa. Second, this information available from another project is supplemented by data collected specifically for this study on the extra-regional contacts of local banks from the five chief Central American cities themselves. These data are simple recordings of links between the local banks and other banks in cities outside the region. These are arrangements between banks that facilitate the movement of capital in and out of the region and, in aggregation, describe a pattern of linkages between Central America and the world city network. These data are also derived from investigation of the web sites of service firms.
Reliance on firm's web sites for most of our data requires justification. These are useful information sources because the sites are designed, in part, as marketing vehicles to indicate the geographical scope of a firm's services. For global firms this means providing information on their office network, for local firms potential customers are shown what arrangements can be made to conduct their affairs outside the region. In both cases we focus on the locational information and not on any other promotional material. Since there appears to be no motive for firms to provide false information on office locations, we take the data derived from these to be reliable indications of inter-city connections.
From these different data we can describe three distinctive ways in which the region is integrated into the world city network.
GLOBAL FIRMS IN LOCAL CITIES
Global service firms vary greatly in their particular global strategies of office location. Smaller firms have obviously to concentrate their resources on less cities but the critical determinant of location policy seems to be the service sector. The greatest contrast is between global accountancy firms that tend to locate in many hundreds of cities and global law firms that are usually found only in a select number of world cities. This means that while the five Central American cities are not world cities, they do have some direct connections into the world city network through those firms that have an extensive office location policy. These are shown in Table 1 where it can be seen that San Jose has most (20) and Managua least (11) of the 100 global service firms which constitute the data. Most of the firms with Central American offices are in accountancy and advertising (the list of firms is given in appendix A). The key point, however, is that, in some sectors at least, Central American cities are part of the office networks that create the world city network. In other words this network is not constituted as an exclusionary club of the major cities but has linkages into regions without world cities.
This direct servicing of Central America through its local cities can be described by measuring the global network connectivity of the region's cities.6 These are derived from the100 x 316 matrix of service values. The global connectivity of each city is computed by first summing the products of its service value with every other city's service value for each firm, and then adding together these 315 sums to produce a global measure of connectivity. The end result of this exercise is 316 measures indicating how well each city is linked, through sharing offices in the same firm's networks, to all other cities in the set. Obviously if a city has no offices for any of the 100 firms the connectivity will be zero (there are 3 such cities in the data). In contrast, London has the highest connectivity measure closely followed by New York indicating that for the vast majority of the firms these two cities are central to their global strategies. (Taylor et al. 2001).
The results from this measurement exercise for the five Central American cities are shown in Table 2. The first point to note is their relatively low rankings (Guatemala City is the highest at #130) verifying the non-existence of world cities in this region. However, these results do provide a measure of the direct servicing of this region through its local cities into the world city network. In the third column the scores are converted into proportions of the overall highest scoring city (London) to provide a relative indication of the connectivities: all five cities have between a fifth and a tenth of London's connectivity. To put these relative levels into an American perspective, New York, Sao Paulo, Mexico City and Miami score 0.98, 0.55, 0.49 and 0.47 respectively.
On the basis of this general assessment of business services our conclusion is that Central America is not strongly connected into the world city network but that it is connected.
LOCAL FIRMS IN GLOBAL CITIES
The obverse of global firms in local cities is the linkages of local firms in global or world cities. Small non-global firms can operate beyond their normal geographical range by forming alliances or having other similar relationships with firms in other regions. This has been a common practice in the banking sector where 'correspondence banks' are designated. Where a local bank has a client doing business in another area where it does not have an office, it will advise and facilitate that financial service being undertaken by its correspondence bank in that area. This is not a formal alliance but indicates a 'partner of choice' for the mutual benefit of both: the correspondent bank gets the extra business, the local bank does not lose its client through failure to provide adequate geographical scope of service. This does not constitute the ideal of a seamless service under one brand but it does provide the opportunity for local firms to service extra-regional business.
We have collected information on the correspondent links of twenty two local Central American banks (these are listed in appendix B). A total of 319 links were found of which 168 were with one city: Miami. That over half (53%) of the correspondent links go to banks in just one world city is quite a remarkable finding, a stark indication of the domination of Miami in the external financial connections of Central America. This is a contemporary manifestation of Foucher's (1988, 121) designation of Miami as the 'capital' of the 'American Mediterranean' in the Reagan era. Far behind in second place, and reinforcing the USA linkage dominance, is the Americas' prime international financial centre New York with 35 (11%) correspondent links. Although there is this concentration of linkages, Central American correspondence banks are to be found across the world in 34 different cities. These are shown in Figure 1 where, as well as emphasising Miami's primacy, there are other interesting features notably the relative importance of Hamburg and Frankfurt, the top two ranking European cities, compared to Madrid with its colonial/language connections and London, Europe's prime international financial centre. The relatively low level of connections with Latin America, excepting the two 'neighbour' cities of Mexico City and Panama City, and the lack of importance of Pacific Asian cities are also noteworthy.
On the basis of this specific financial link, our conclusion is that Central America has widespread indirect links into the world city network but that these are hugely dominated by its connections to Miami.
GLOBAL FIRMS AND REGIONAL CITIES
Miami is not a major world city according to most lists of such cities in the literature (Beaverstock et al., 1999). There is, however, one important area where Miami does turn out to be very highly ranked. When global service firms choose to organise their business through designated 'regional offices' that function to direct a firm's business in a given world regional zone, Miami is found to be one of five prime 'regional world cities' along with Hong Kong and Singapore plus London and New York which have regional responsibilities alongside their truly global functions (Taylor, 2000). This is another form of indirect linkage, a top-down version. Information on regional offices is available for the 100 global service firms used previously. However, in this scheme of things, Central America is not a large enough region to constitute a zone with a regional office. Rather there are regional offices for Latin America, which, of course, can service the Central American 'sub-region' as required.
Not all firms use a regional framework for their global organisation but nearly one fifth (19) of our 100 global service firms do have designated Latin American offices (these firms are listed in appendix C). There are a variety of titles used for these 'regional offices' - regional headquarters, regional executive office, senior partner's office for the region, zonal administrative office, office of the group head in the region - but they all imply the management of servicing Latin America including Central America. From whence is this directing coming from? Interestingly, Table 3 confirms the dominance of Miami despite the fact that we are now covering a much larger world region. Nevertheless with 11 regional offices Miami has 58% of the regional responsibilities within the Latin America zone. This is even more remarkable because the city is not in Latin America as usually defined. It is a sort of 'extra-mural capital' of Latin America.7 Of the other main Latin American world cities, Sao Paulo is Miami's only rival but is a long way behind and Mexico City does not feature at all as a regional centre.
On the basis of this regional office information, our conclusion is that Miami is confirmed as a key link, this time hierarchical, for Central America (as part of Latin America) integration into the world city network.
The objectives of this paper were cast in two forms and therefore we provide two brief conclusions from these results. Our basic claim is that from both perspectives we are opening up new research agendas which we address briefly below.
From a world city perspective we have shown relations between cities that go well beyond the usual cast of major metropolitan centres. There is a strong tendency for world city studies to focus on the most important nodes in the network to the relative neglect of much of the rest of the world. By focusing on a region 'beyond world cities' we have instigated research on a new category of inter-city relations involving minor cities in a small region. From our empirical evidence it is clear that the world city network is not a simple bounded web of relations involving major cities only. At the fuzzy edge of the network of major cities there are numerous tree graphs, single links that connection smaller cities into the network. The implication is that the world city network is truly global.
From a Central American perspective it is clear that the transnational model built upon Neoliberal domestic reform is complemented by a new form of external orientation, the 'global circuits' that Robinson (1997a: 131) refers to. In specifying some of these connections we have shown that in finance and other business services the relatively small cities of Central America are linked into a new and evolving world city network that is servicing global capital. It would seem that there are no 'backwaters' in this process, regions are integrated in relation to the size of their local market. Central America may be a small region economically but it has changed appreciably as part of the wider process of economic globalization and its new spaces of flows. The implication is that Central America is truly global.
These two perspectives come together in our findings on Miami. This city is clearly crucial to Central America's connection into the world city network. As a world city it has often been portrayed as unusual - the 'most foreign city' in the USA (Nijman, 1997: 164), the city that 'breaks the rules' (Nijman, 1996), a contingent political (CIA) creation (Grosfoguel, 1995), with totally distinctive connections (Taylor and Walker, 2001) - but this is the first time the city's unique role as a gateway city8, a hinge between the two Americas, Anglo- and Latin, has been empirically illustrated as part of a global space of flows. While Miami attends to Central America and connects the region to a wider world, in its own small way Central America attends to Miami and helps create a major regional world city out of what otherwise is a small to medium-sized world city.
Initially we presented this study as a preliminary exploration of a particular segment of global integration. The next stages of the research are twofold. First, from the world city network perspective there is a need to explore other regions that are 'beyond world cities' to compile more information about how the world city network operates globally. Second, we need to investigate further through fieldwork the linkages uncovered in this research: how, by whom and for what are the links operated and in what ways do they relate to other sectors of the regional economy. More generally there is the fundamental geographical issue of how the global space of flows, as represented by the world city network, is interacting with spaces of places such as territorial states, both large and small, and world-systems zones, both core and periphery, in the reproduction of the world economy under conditions of contemporary globalization. Clearly, in this paper we have presented only a small empirical tip of a huge research iceberg.
David Walker computed the connectivity results in Table 2 for which we thank him. We acknowledge also the support of the ERSC in funding the project "World City Network Formation in a Space of Flows" from which this paper derives.
Abell J (1999) 'The Neoliberal World Order: The View from the Highlands of Guatemala' NACLA Report on the Americas 33 (1), 37-41
Beaverstock J V, Smith R J and Taylor P J (1999) 'A roster of world cities' Cities 16, 445-58
Bendana A (1999) 'Nicaragua's Structural Hurricane' NACLA Report on the Americas 33 (2), 16-23
Brown E (1994) Agroexports and Sandinismo: The Political Economy of Social Transformation in Nicaragua (1979-1990) Unpublished PhD Dissertation, Faculty of Social Sciences, University of Edinburgh
Brown E (2000) 'Still their Backyard? The US and post-Mitch development strategies in Nicaragua' Political Geography 19, 543-572
Borjas P (1989) 'Perspectives on the Central American Crisis: Reactionary Despotism or monopoly capital?' Capital and Class 39, 51-81
Bulmer-Thomas V (1991) A Long-run Model of Development for Central America' University of London, Institute of Latin American Studies Research papers no. 27
Camacho D and Rojas M (eds) (1984) La Crisis Centroamericana (Editorial Universitaria Centroamericana/Facultad Latinoamericana de Ciencias Sociales, San Jose)
Castells, M. (1996) The Rise of Network Society (Blackwell, Oxford)
CEPAL (1986) 'Central America: Bases for a Reactivation and Development Policy' CEPAL Review 28, 11-48
Dauner I (1998) 'Liberalizacion Financiera y Segmentacion de Mercados' Encuentro 45, 3-17
Dijkstra A (1999) 'Technocracy Questioned: Assessing Economic Stabilization in Nicaragua' Bulletin of Latin American Research 18 (3), 295-310
Dore E and Weeks J (1992) 'Up from Feudalism' NACLA Report on the Americas 26(3), 38-44
Dunkerley J (1988) Power in the Isthmus: A Political History of Modern Central America (Verso, London)
Evans T, Castro C and Jones J (1995) Structural Adjustment and the Public Sector in Central America and the Caribbean (CRIES, Managua)
Evans T (1998) Liberalizacion Financiera y Capital Bacario en America Central (CRIES, Managua)
Friedmann, J (1986) 'The world city hypothesis' Development and Change 17(1), 69-84
Foucher, M. (1987) 'Geopolitical approaches to the "Mediterranean Basin" of America' in Girot, P. and Kofman, E. (eds) International Geopolitical Analysis. A Selection from Herodote, (Croom Helm, London), 106-31
Grosfoguel, R (1995) 'Global logics in the Caribbean city system: the case of Miami' in Knox P L and Taylor P J (eds) World Cities in a World-System, (Cambridge University Press, Cambridge), 156-70
IMF/IDA (International Monetary Fund/International Development Association) (1999) 'Nicaragua: Preliminary Document on the Initiative for Heavily Indebted Poor Countries (HIPC)' August 23, IMF/IDA, Washington
Judson F (1993) 'The Making of Central American national agendas under adjustment and restructuring' Labour, Capital and Society, 26 (2), 148-80
Kaimowitz D (1990) 'The 'Political' Economies of Central America: Foreign Aid and Labour Remittances' Development and Change 2, 637-655
La Feber W (1993) Inevitable Revolutions: The United States in Central America (Norton, New York)
Lindenberg M (1988) 'Central America: Crisis and Economic Strategy 1930-1985, Lessons from History' The Journal of Developing Areas, 22(1), 155-178
Lo, F-C and Yeung, Y-M (eds) (1996) Emerging World Cities in Pacific Asia (United Nations University Press, Tokyo)
Lo, F-C and Yeung, Y-M (eds) (1998) Globalization and the World of Large Cities (United Nations University Press, Tokyo)
MSUSPSI (Minority Staff of the US Senate Permanent Subcommittee on Investigations) (2001) Correspondant Banking: A Gateway for Money Laundering (US Government Printing Office, Washington)
Nijman J (1996) 'Breaking the rules: Miami in the urban hierarchy' Urban Geography 17, 5-22
Nijman J (1996) 'Globalization to a Latin beat: the Miami growth machine' Annals, American Academy of Political and Social Sciences 551, 163-76
Orozco M, de la Garza A and Baraona M (1997) Immigracion y Remesas Familiares' Cuadernos de Ciencias Sociales de FLACSO 98, March 1997, FLACSO: San Jose.
Pelupessy W (ed) (1991) Perspectives on the Agro-export economy in Central America (Macmillan, Basingstoke)
Pelupessey W and Weeks J (eds) (1993) Economic Maladjustment in Central America (Macmillan, Basingstoke)
Robinson W (1997a) 'Maldevelopment in Central America: A study on Globalization and Social Change' Pensamiento Propio 5 (nueva epoca), 131-164
Robinson W (1997b) 'Nicaragua and the World: A Globalization Perspective' in Walker T (ed) (1997) Nicaragua without Illusions: Regime Transition and Structural Adjustment in the 1990 (Scholarly Resources, Wilmington, Delaware), 23-42
Rodas-Martini P (nd) 'Centroamerica: Para afrontar con exito la globalizacion del siglo XXI' CA 2020: Documento de trabajo # 1 (http://ca2020.fiu.edu/Themes/Pablo_Rodas/rodas.htm)
Sassen, S (1991) The Global City (Princeton University Press, Princeton, NJ)
Sassen, S. (1994) Cities in a World Economy (Pine Forge, Thousand Oaks, CA)
Short J R, Breitbach C, Buckman S and Essex J (2000) 'From World Cities to Gateway Cities', City 4, 317-340
Siddiqui K (1998) 'Agricultural exports, poverty and ecological crisis' - Case study of central American countries' Economic and Political Weekly 33 (39), A128-A136
Sojo C (1999) Democracias con Fracturas: Governabilidad, Reforma Economica y Transicion en Centroamerica (Facultad Latinoamericana de Ciencias Sociales, San Jose)
Taylor P J (2000) 'World cities and territorial states under conditions of contemporary globalization' Political Geography 19, 5-32
Taylor, P. J. (2001) 'Specification of the world city network' Geographical Analysis 33, 181-94
Taylor P J and Walker D R F (2001) 'World cities: a first multivariate analysis of their service complexes' Urban Studies 38, 23-47
Taylor P J, Catalano G and Walker D R F (2001) 'Measurement of the world city network', GaWC Research Bulletin 43
Torres Rivas E (1969) Interpretacion del Desarrollo Social Centroamericano (Editorial PLA, America Nueva: Chile)
Torres-Rivas E (1989) Central America: Repression and Resistance. (Westview, Boulder)
Torres-Rivas E (1993) History and Society in Central America (University of Texas Press, Austin)
Vilas C (1995) Between Earthquakes and Volcanoes: Market, State and the Revolutions in Central America (Monthly Review, New York)
Weeks J (1985) The Economies of Central America ( Holmes and Meier, New York)
Weaver F (1994) Inside the Volcano (Westview, Boulder)
Weeks J (1999) 'Trade Liberalisation, Market Deregulation and Agricultural Performance in Central America' Journal of Development Studies 35 (5), 48-75
Williams R (1986) Export Agriculture and the Crisis in Central America (The University of North Carolina Press, Chapel Hill)
Wilmore L (1997) 'Export Promotion Policies in Central America' CEPAL Review 62,
Williams (1994) States and Social Evolution: Coffee and the Rise of National Governments in Central America (The University of North Carolina Press, Chapel Hill)
1. For more detail on the key significance of the Agro-export sector in Central American economic development see Williams (1994) & (1986), Siddiqui (1998), Pelupessy (1991), Lindenberg (1988), Bulmer-Thomas (1991), Pelupessy and Weeks (1993) and CEPAL (1986).
2. The social and political repercussions of the agroexport model in Central America were first coherently analyzed by Edelberto Torres Rivas in his seminal work Interpretacion del Desarollo Social Centroamericano (1969) - much later published in English as Torres Rivas (1993). Other useful interpretations include Camacho and Rojas (1984), Weeks (1985), Dore and Weeks (1992), Weaver (1994), Borjas (1989), Dunkerley (1988), Vilas (1995), Torres Rivas (1989) and La Feber (1993).
3. For other treatments of the Neoliberal reforms adopted across Central America over the past fifteen years see Brown (2000), Bendana (1999), Weeks (1999), Sojo (1999), Robinson (1997a), Evans et al (1995), Dauner (1998), Dijkstra (1999), Abell (1999) and Judson (1993).
4. There have, of course, been several studies of world cities in the 'third world' (see, for instance, in Lo and Yeung (1996, 1998) and 'semi-peripheral' cities were included in Friedmann's (1986) initial ordering of world cities but no studies that we are aware of that look at the world city network in regions without a world city.
5. Global service firms are defined by having offices in 15 or more different cities and at least one office in each of western Europe, Pacific Asia and the USA. The firms in this study are in the following sectors: 18 in accountancy, 15 in advertising, 23 in banking/finance, 16 in law, 11 in insurance, and 17 in management consultancy.
6. The following is based upon the detailed description of the procedure found in Taylor et al. (2001).
7. Actually, one firm lists Miami as a Latin American country!
8. Unlike Short et al. (2000) we do not pose gateway city as an alternative to world city but treat it as a role that some world cities perform.
Table 1: Sector representation of global firms in Central American cities
AC = accountancy; AD = advertising; BF = banking/finance; IN = insurance; LW = law; MC = management consultancy
Table 2: The global connectivity of Central American cities
* these are proportions of the highest recorded connectivity, that of London.
Table 3: Regional offices for Latin America
APPENDIX A: Global firms with Central American offices (see Table 1)
APPENDIX B: Central American banks used in the survey of correspodent linkages (see Figure 1)
APPENDIX C: Global firms with Latin American regional offices
Figure 1: Central American connections with the world city network: correspondence banking contacts (for the list of banks included, see Appendix B)
Edited and posted on the web on 2nd July 2001; last update 5th February 2002
Note: This Research Bulletin has been published in Area, 34 (2), (2002), 139-148