GaWC Research Bulletin 22

Geographers, and other social scientists, have been debating the virtues of globalization since the 1990s (Allen and Hamnett, 1995; Amin and Thrift, 1994; Castells, 1996; Cox, 1997; Dicken, 1998; Featherstone, 1990; Hirst and Thompson, 1996; Storper, 1997). For example, Amin and Thrift (1997) point to five globalizing tendencies: (1) the globalization of money and financial capital; (2) the importance of knowledge-structures as a factor of production; (3) the internationalisation of technology; (4) transnational oligopolies; and (5) the rise of transnational diplomacy between firms and states. But here, like in other globalization texts (Lee and Wills, 1997), studies of migration have remained transparent in the literature.

It has been left to population geographers to study the links between globalization and migration. Recently, we have been able to draw upon a rich vein of research investigating globalization tendencies in the context of diasporas, transnational communities, gender and international migration (Castles and Miller, 1993; Cohen, 1997; Conway and Cohen 1997; Findlay 1988; Gould and Findlay, 1994; Hardhill and MacDonald, 1998; Wong, 1997). In the context of highly-skilled migration, but beyond discussions of brain drain (Hague and Kim, 1995), an extremely fruitful analysis of globalization and migration has come from studies of 'transient' skilled migration within transnational corporations (TNCs) (Appleyard, 1991; Salt and Findlay, 1989), and especially the work of John Salt (Koser and Salt, 1997; Salt, 1988; 1992). But, transient migration still remains an 'invisible' phenomenon, in both an empirical and theoretical sense (Findlay, 1996).

In this paper, we wish to build upon the work of those geographers, like Findlay and Salt, who are already examining migration as a globalization tendency. Accordingly, the major aim of this paper is to investigate the symbiotic relationships that exists between globalization, TNCs and transient professional migration, but in the context of global city international financial centres (IFCs). The rest of this paper is divided into four inter-linking parts. After this introduction, part two develops a conceptual framework, which discusses transient professional migration as an important globalization tendency within the global city IFC. Part three, reviews a series of qualitative methods, which have extended our knowledge of transients in globalization studies. Part four, reports an analysis of professional and managerial migration trends from the International Passenger Survey, and moreover, introduces new empirical work, which investigates transient professional migration within TNC banks, between 1993-1999. Finally, in part five, several conclusions are reported.

NEGOTIATING GLOBALIZATION, TRANSNATIONAL CORPORATIONS AND GLOBAL CITY FINANCIAL CENTRES IN TRANSIENT MIGRATION STUDIES

As we shall discuss now, the study of migration, as a globalization process, has long been associated with global city (re)production and the organizational strategies of TNCs.

Globalization, Migration and Global Cities

Migration has been central to the global city debate. As Friedmann and Wolff (1982, 322-23) and Friedmann (1986, 75), commented respectively:

"Transnational elites are the dominant class in the world city, and the city is arranged to cater to their lifestyle and occupational necessities".

"The world city hypothesis is about the spatial organization of the new international division of labour ... world cities are points of destination for large numbers of both domestic and/or international migrants".

It has been Saskia Sassen, however, who has taken the lead in conceptualising the role of globalization and migrant labour in the global city. In (1998) The Mobility of Capital and Labour, (1991) The Global City and (1994) Cities in a World Economy, and numerous journal articles, she has emphasised the rise of the low-waged immigrant sector, supporting the professional elite, as being crucial to global city (re)production (Sassen-Koob,1986; Sassen, 1987). But, in all her work, Sassen consistently downplays the role of transient professional migrants in global city (re)production: and, thus, continues to reinforce their invisibility with both the global city context, and world-system.

Building upon these seminal works, others have begun to investigate the different ways in which migration as a globalization process is (re)producing the global city. Here, of particular importance, has been the collective work of: the Institute of British Geographers Limited Life Working Party on Skilled International Migration (Findlay and Gould, 1989); Beaverstock on world cities and advanced producer services (Beaverstock, 1994); Findlay, Jones, Jowett, Li and Skeldon on Hong Kong (Findlay and Li, 1997; Li, et al 1998); and studies of gender and immigration (Kofman, 1996). With very few exceptions, all of these studies have theorised migration as an integral globalization process, which has produced new geographies of migration, with respect to both process and pattern.

Globalization, Transnational Corporations and Transient Migration

The TNC provides an enabling environment for the (re)production of transient professional migration in the world-system. TNCs use transient professional migrants in order to bring human capital and intellectual knowledge to their foreign operations (Beaverstock, 1991; Cormode, 1994; Perkins, 1997; Salt, 1984; 1988; 1992). Equally, such professionals are encouraged to enhanced their career paths within such organizations by spending time working abroad (Beaverstock, 1996b; Salt, 1988). Further, given the disproportionate location of TNCs and specialised high-order service functions within global cities, substantial flows of transients occur between them (Economist, 1998). But these flows are very difficult to quantify given the poverty of disaggregated data from official sources and ICTs (Salt et al, 1994). Moreover, as globalization processes are continuing to concentrate and intensify specialised service TNC functions within the financial spaces of global cities (Sassen, 1995), any analysis of transient professional migration, must investigate them within the context of IFCs.

If we are to explore transient professional migration between IFCs, we must go further than Sassen, and the work on TNCs. We should begin to link the globalization tendencies of professional migration with theories accounting for financial geographies within IFCs (Leyshon and Thrift 1997). In the following section, we discuss the symbiotic relationship that exists between globalization and transient migration in IFCs by considering two very important 'driving forces': producer service TNCs; and migrants' embedded geographies - their role in the accumulation of 'cultural capital'.

Transnational Corporations and Transient Migration in International Financial Centres

Transient professional migration remains an 'invisible' facet of globalization processes within IFCs (Beaverstock, 1996a). This is surprising, when we consider that London, New York, Tokyo and Singapore, have become 'global complexes' not only of economic power, but also social and cultural influences in the world economy (McDowell, 1997; Sassen, 1995; Thrift, 1994). Important constituents of this 'global complex' are those transient professional international workers who move within advanced producer service office networks in IFCs. Recent findings from Beaverstock's (1996a-c) analysis of transient professional migration in accountancy and investment banks in the late 1980s and early 1990s, revealed that TNCs continued to post staff to IFCs, despite improvements in information technology, high costs and real-time financial media networks. Staff were posted to IFCs, because they had, or could obtain, the specific knowledge, expertise, and skills required to ensure the efficient operation of the financial system and the global reach of the TNC. In accountancy and banking, transient professionals performed important 'global face-time' processes, between firm and client, and international markets. However, an important function of transient professionals is also to network, and accumulate 'cultural capital' (McDowell, 1997), as a part of their everyday expatriate experiences.

Transient Migrants as Cultural Capital

Couched within the new economic geography (Lee and Wills 1997), there has been a myriad of work which discusses 'embedment', knowledge, expertise and networks, as 'global' processes which accumulate cultural capital within IFCs (Amin and Thrift, 1997; Leyshon and Thrift 1997; McDowell 1997). In this work, the emphasis has been focused towards investigating the organizational cultures of professional staff, their knowledge, and the production/circulation of that knowledge. In the City of London, they argue that such knowledge and reflexivity is achieved through an individual's: involvement in business and social networks; cultural life experiences; gender relations; wealth; 'meeting places'; and, reaction/involvement in the day-to-day financial atmosphere, tacitly couched in story making and rumour. Equally, we would like to suggest that the performance of a financial TNC is also very much linked to the success and speed with which their transient migrants, accumulate/circulate knowledge, expertise and 'intelligence' in, and out of, the institutional workplace. As Thrift (1994, 336) argues, an important element of the City of London's 'global corporate network' is its 'constant through-flow of workers' from other IFCs. Or to put it another way, the competitiveness of a financial TNC in New York City, is just as dependent upon the success of its transient professional migrants in creating cultural capital, than creating capital from conventional financial transactions (Beaverstock, 1996b), as a reading of Nick Leeson's (1996) Rogue Trader highlights. Moreover, the cultural capital obtained from overseas postings is highly prized amongst individuals. As one foreign banker states:

"From the late 1960s or early 1970s [..] it is possible to trace people sent to run the London branch for two or three years who ended up a few years later being the chairmen . of the top US investment banks." (The Financial Times, 1997, 25).

Drawing upon the major contextual material discussed above, in the rest of this paper, we wish to explore the linkages that bind together globalization and migration, as process, with reference to transient professional migration geographies within TNC commercial banks. But, first, we briefly qualify our methodology, and reflect upon the approach of recent research investigating globalization, TNCs transient professional migration and global cities.

RESEARCHING GLOBALIZATION, TRANSNATIONAL CORPORATIONS AND FINANCIAL CENTRES IN TRANSIENT MIGRATION STUDIES

Since the late 1980s, a small group of geographers have begun to focus on globalization, TNCs, migration and the global city (as listed in Table 1). But, moreover, these writers have become instrumental in bringing qualitative approaches into the study of migration. Table 1 reveals three key methodological tendencies in their work. First, many of the early writings (late 1980s) were based upon the analysis of official aggregated Census of Population or Employment statistics (especially Sassen's work) and, or firm ICT data (see Salt, 1988). Second, in the early 1990s, with exceptions (Salt, 1992), research was often executed by postal questionnaire surveys and associated statistical analysis. The rationale for using these quantitative techniques was to collect primary material, which would indicate 'representative' trends in migration data. Third, from the mid-1990s, and influenced by other geographical practices, researchers have been adopting mixed-method approaches to provide a biographical approach to understanding migration (Halfacree and Boyle 1993), with particular emphasis placed upon semi-structured interviewing, textual analysis and life histories. It has been the mixed-method approach which has provided much 'added value' and extended our theoretical (process) and empirical (pattern) understandings of globalization, TNCs and migration tendencies, especially in the global city.

The merits of using mixed-methods in migration research have now been recognised (McKendrick, 1996) and in the following case study, we wish to make a positive contribution to such research, in two ways. First, we illustrate how transient professional migration has become an integral part of globalization by drawing upon official statistics and case study research undertaken in one very important globalization 'driver' within IFCs: TNC banks. Second, we extend the methodological virtues of using taped semi-structured interviews, as a mechanism with which to explore a richness of detail and depth in findings, and trace processes, which could never be found by using other methods.

GLOBALIZATION AND GLOBAL CITY TRANSIENT MIGRATION: EVIDENCE FROM THE BANKING SECTOR IN THE 1990s

First, we will review the major trends of international professional and managerial migration, as reported by the United Kingdom's (UK) Office for National Statistics (ONS).

The International Passenger Survey

Official data sources reporting international migration are collected from the International Passenger Survey (IPS). The IPS is a sample survey, based upon structured interviews with passengers who depart/arrive at the UK's major airports, seaports and Channel Tunnel. As a sample survey, a major limitation of IPS data are the standard errors, especially when analyzing disaggregated data (Findlay and Gould, 1989). In 1997, the 93,200 and 85,500 professional and managerial migrants who entered and left the UK, were estimated from 565 and 269 contacts, which produced standard errors of 7.5 and 9.9 per cent, respectively (ONS, 1999, 35). Thus, any analysis using IPS data cells, must be undertaken, using Findlay and Garrick's (1990, 181) phrase, with a "cautionary note."

An analysis of ONS professional and managerial migration reveals four major trends over the last ten years (1988-1997) (Table 2, Figure 1a-e). First, if we investigate net migration from 1988-1997, the data shows that immigration surpassed emigration on six occasions in the UK (1988, 1991-1993 and 1994-1997) and seven in England and Wales (with the addition of 1991) (Table 2, Figure 1a-b, Appendix 1). Moreover, the time-series data shows that England and Wales continually accounted for the highest proportions of immigrant and emigrant professional and managerial migrants (Figure 1c). Second, if we examine the gender breakdown of professional and managerial migrants, more men left and entered the UK than women, between 1988 and 1997 (Figure 1a). On average men accounted for 63 and 62 per cents of immigrants and emigrants, respectively, over the ten year period, which increased slightly to 65% for immigration in the 25-44 year age group (Appendix 1).

Third, an analysis of professional and managerial migration by age, reveals the striking importance of the 25-44 group, for both immigration and emigration, between 1988-1997 (Appendix 1, Figure 1d). When we calculated the proportion of the 24-44 age group as a percentage of all ages, it averaged a staggering 73 per cent for both immigration and emigration, between 1988-1997. This, however, was not unexpected, as a high proportion of professional and managerial labour enter the market late because of economic inactivity during higher education. Fourth, an investigation into the citizenship of professional and managerial migration shows the domination of British nationals in both immigration and emigration (Appendix 2, Figure 1e), between 1988-1997. Moreover, a more detailed analysis of net professional and managerial citizenship data indicated that British emigration exceeded immigration every year (except 1994) and that All Commonwealth and Other Foreign immigration exceeded emigration over the ten years (except 1992 for Other Foreign).

Unfortunately, we cannot access published professional and managerial migration data, which reports length of stay abroad, country of next residence by United Kingdom government region of origin/destination, and main reasons for migration. The ONS do not disaggregate such IPS data because of the magnitude of sampling errors associated with such small data cells. Thus, in order to investigate transient migration in more detail, data has to be collected from other sources. In recent years, such pragmatic research has led to the collection of data from TNCs, as extensively undertaken by writers such as Beaverstock (1994) and Salt (1992).

Case Study Evidence from the Banking Sector

Four important reasons account for the selection of commercial banking, as a medium to investigate the relationship that exists between globalization and migration, in global city IFCs. First, banking is one of the major global forces (re)producing economic power in IFCs (Daniels, 1993). Second, banking is a knowledge industry, which sells ephemeral products, where clients essentially buy the expertise of selected individuals employed by the firm (Laulajainen, 1998). Third, the cultural and social geographies of work in IFCs are constructed from the 'polarised' production and consumption patterns of the advanced producer service workforce (Sassen, 1991). Fourth, banking has also been long associated with facilitating transient professional migration between IFCs (Beaverstock, 1994).

For example, the City of London had 565 foreign banks in 1998, which employed on average between 35,000 - 40,000 staff (Corporation of London, 1995; Reed, 1998). In contrast, what is less known is that these foreign banks bring with them a multitude of transnational workers into the City each year (Beaverstock and Smith, 1996). These workers, however, are extremely difficult to quantify. They remain too disaggregated to be identified in official statistics like the Labour Force Survey. Despite such difficulties, we must not underestimate the magnitude of transnational labour in the City of London. As the Economist (1998, 31) reports:

"Goldman Sachs now number 66 different nationalities amongst its 2,300 employees, whilst 35% of the workforce of J.P.Morgan are non-British citizens . [1,200 foreign staff] .... Take out the 'backroom' and secretarial staff from the equation . [who are local nationals] . and the international nature of these banks becomes even more striking."

The ratios of foreign staff presented by the Economist (1998) in Goldman Sachs and J.P.Morgan are not unrepresentative examples of migrant labour found in the larger foreign banks present in the City. As Reed (1998, 43), comments " . thousands of financial professionals from the US, Asia and Europe . are flocking to the City . led by US banks and investment houses". In the remainder of this section of the paper, we will report original research findings, which highlight TNC banks' transient professional migration practices in 1993 and 1999, and illustrate the relationship that exists between globalization and migration in IFCs. But, first, it is important to discuss the research methodology.

Methodology

In April 1993, the top ten United Kingdom (UK) banks (ranked by Market Capital $m) were selected for interview, drawn from The Banker's survey of the 'Top 1,000 by country': Barclays Bank; National Westminster Bank; Abbey National; Lloyds Bank; HSBC Holdings Plc; TSB Group; Royal Bank of Scotland; Bank of Scotland; and Standard Chartered (Blanden and Timewell, 1992). Four banks were successfully interviewed, representing a 40 per cent response rate. The same selection procedure was adopted for the April 1999 survey. This research is still ongoing, and to date (June 1999), the response rate has been thirty per cent, but we have not been able to interview one 1993 respondent.

For both surveys, data was collected through 1-2 hour taped semi-structured interviews with the banks' Director of Human Resources, at the London headquarters. The rationale for interviewing expatriate human resources specialists was that they were generally the line managers responsible for planning, organizing and executing expatriate postings, within the bank's international office network, who also had the authority to release confidential information. For both periods, the interview schedules remained constant, seeking information on the banks': international structure; magnitudes and geographies of labour migration; and corporate migration policy. After interview, transcriptions were returned to the interviewees for editing. This reinforced the need for confidentiality and trust. In order to retain the confidentiality of the 1999 interviews, specific banks are replaced with pseudonyms: Northbank; Southbank; Eastbank and Westbank.

Corporate Migration

In order to highlight the richness of the data obtained from using interviews as a methodological tool, we wish to proceed by presenting the detailed studies of each bank in turn, followed by a discussion. In the discussion, we will draw together the salient characteristics of these data and IPS findings, and provide a critique of the methodology used.

Northbank. In 1993, Northbank had 200 International Managers (IMs), working in cross-border activities throughout its 750 international offices (Table 3). They also had 150 Secondments, whose length of stay abroad was between 2-3 years. Secondments were distributed unevenly throughout the bank's offices, but IMs were concentrated in Asia-Pacific (e.g. Singapore, Hong Kong, Shanghai), African (e.g. Nairobi) and European regions of the bank (e.g. Paris, Madrid, Frankfurt, Milan, Zurich). In addition to IMs and Secondments, the Expatriate Manager (EM) suggested that "many cross-secondments . [were] . taking place at any one time within the Group . to develop high calibre local managers, by means of providing them with international opportunities." Thus, within Northbank, migration was a central globalization strategy. In all cases, staff were sent abroad to perform corporate banking, information technology, risk and treasury functions, all of which had been created by skill shortages in foreign offices. In addition, IMs circulated within the bank's global network to manage offices. In 1993, all foreign workers entering London were treasury, credit or personal bankers, originating from a wide range of foreign offices. With respect to gender, the majority of IMs and Secondments were male, however, the bank did point out that they were unable to send women to certain Middle East offices. Finally, looking towards the future, the EM predicted that the number of IMs would decline (by as much as 50%), as more local offices would resource themselves, however, with respect to Secondments, he reported that there would be very little change, except "a more multi-cultural mix of secondees".

Northbank would not grant an interview in 1999. But, the following data is derived from The Financial Times and the banks WWW site (sources undisclosed to protect confidentiality). At the end of the 1990s, transient professional migration still remains an important globalization strategy of the bank. It had increased its IM and Secondments to 420, 70 more than in 1993 (The Financial Times). The raison d'être of the IM has remained the same: preaching the banks global culture and providing technical knowledge and expertise, when critical skill shortages exist. Looking towards the next millennium, the bank is planning to use many more Secondments, involving all nationalities, in order to become "transnational" in scope, and offer a "seamless" global service operating across national-borders (Northbank, Head of Human Resources, quoted in The Financial Times; WWW site, accessed 20 January 1999).

Southbank. In 1993, it "had approximately 600 International Officers (IOs) who continually moved on cross-border assignments between its 3,000 offices in 68 countries" (Expatriate Manager [EM]). These personnel were an integral element of the banks' global practices, and were employed as branch managers or very senior staff, specialising in "all technical functions", throughout the global office network (EM). During that time, the 40 Secondments, went for periods of 2-3 years, and fulfilled the functions of private, treasury, capital markets and information technology, as required to balance skill shortages created in hosting offices. With respect to recruitment, the "business line manger . decides on the personnel to be sent overseas" and, more importantly, "staff should already have the necessary skills before they take up an overseas assignment" (EM). But, differences did exist between function and length of stay abroad:

"Banking, Treasury, Capital Markets, IT, Finance and Audit are the specialist functions personnel fulfil when on overseas assignments. Different functions require different modes/speeds of performance - e.g. someone in sales may take longer to become productive, since the development of relationships is important to the job . Specialists in capital markets and treasury can go on overseas assignments for relatively short time periods due to the homogeneous environments they will be working in" (EM)

All moves from the UK were through London, the 'sanctioning environment'. London was regarded as the "bank's centre of excellence and this is where many of the high flyers are located, and the skills which are always in demand" (EM). In 1993, the major destinations for assignments were: New York City; Paris; Frankfurt; and, Hong Kong. The bank received staff from its Group headquarters, outside of the UK, and their function was to perform very senior treasury, corporate finance and merchant banking activities. With respect to gender, the majority of IO's and Secondments were male. As to the future:

"we have no substantial changes to make to the numbers of staff going abroad over the next five years . and as yet there is no recruitment drive towards employing European bankers . moreover, as local offices foot the bill for expatriate staff, there is now more consideration for using . [them] . in overseas locations" (EM).

In 1999, Southbank maintained its IO's, on permanent cross-border postings, and are used exclusively to "meet the business resourcing needs of the Group" (WWW site). IO's are now also recruited directly from graduate entry. They receive a very structured career path which starts with the Executive Trainee Development Programme (at the Group Management Training College), and after successful completion, is followed by the five year Junior Officer Development Programme (JODP), which involves a posting to one of the Groups major 5,500 offices, in 79 countries. The distribution of the 50 IO's currently on JODP's breakdown as: 72% posted to Hong Kong; 15% posted to New York and Toronto; and the remainder to Europe. After successful completion, staff become fully-fledged IOs, and are then posted every 2-3 years between offices in the Group (WWW site, accessed 20 January 1999).

An interview was undertaken with the Director of Human Resources in April 1999 (DHR). In contrast to 1993, much more information was given on IOs and Secondments, but, very little was stated on immigration, gender and training. With respect to Secondments, they were:

"essentially technicians, who supplied their specific skills on demand . and are very much vacancy driven. Either because we can't get an international officer or because we need a particular set of skills which are unlikely to be found in the international officer . secondees will also be expected to rise to senior management with their particular business, if not to run the business itself" (DHR).

In contrast, the IO embodied the culture of the bank at a global scale:

"they are recruited to be permanently mobile . to get a broad range of functional and geographic experiences across the group because ultimately it is expected that they will have the potential to rise to the higher levels of management within the group and to do that you need to have a view across the group. Both our Chairman and our Chief Executive were 30 years as IO's before they took up their positions" (DHR).

The DHR stressed that IOs were maintained because:

" . at a metaphysical level it essentially involves the culture of the Group . all I'm saying is that what they bring to the role is not primarily technical skill . first and foremost we bring the commitment to the group rather than to the local business: our agenda is the groups' benefit. We bring the network."

Of significant important here, is this last phrase, "we bring the network." Clearly Southbank used IOs as an integral organization policy, to tacitly link office with office, and office with client. But, two interesting trends appear in the migration data (Table 3). First, there has been a 38 per cent net increase in migration (+240 persons), due to the rapid increase in Secondments from 40 to 500, (+1250%), "who go on technical project led work" (DHR). Second, the number of IOs had decreased by a magnitude of 37 per cent (-220), partly attributed to rationalisation of cost structures and an "equalling out" of IOs between Hong Kong, London and New York, and hardship centres (i.e. less IOs circulating between the developed economies).

Eastbank. In 1993, they sent staff abroad in two operational capacities. First, they had 100 Assignments who worked outside of the UK on a third-country-national basis. Assignment staff were posted overseas to become branch managers and "have a 100% mobility clause . specifically to work abroad" (Head of Overseas Personnel [HOP]). But, if the move became more permanent, such staff would become employed as a local national in the host country. Second, they had 175 Secondments, who would return to London after their tour of duty. For Assignments and Secondments, time scales varied:

"the more senior the post, the longer the time spent abroad .. A country manger expects to be on overseas assignment for five years. Personnel on Management Development Programmes will go overseas for a period of 12 months" (HOP).

Whilst Eastbank had a Management Development Programme for all staff in their mid-twenties, staff were primarily sent overseas for business reasons, "to bring with them the bank's ethos, as well as in-depth knowledge of the organizational structure" (HOP). This was especially the case for Assignments. But:

"personnel are seconded overseas . when no local skill is available for hire . for added value, for example to . handle the business of a UK multinational client . for Management Development Programmes . and in the vast majority of cases when an individual is sent overseas it is to undertake a specific role" (HOP).

In 1993, incoming foreign staff originated from offices in eleven different countries, including the USA, France and Kenya, and they entered the UK for training (to London), or technical banking functions (to London and other offices, e.g. Bristol). Turning to gender, the majority of Assignments "tended to be male, although the bank is an equal opportunities employer" (HOP). The HOP concluded, "the number of expatriate personnel will decrease within the next five years" partly attributed to "using local skills to offset the high cost of expatriate location".

In April 1999, a Human Resources Director (HRD) was interviewed at a subsidiary of Eastbank, who was responsible for expatriate resources throughout the Eastbank Group. The interviewee was not the same person as interviewed in 1993. Six years on, the bank had retained its Assignment and Secondment staff, and was still using Management Development Programmes. A major difference in policy from 1993, however, was the new ad-hoc category involving information technology migrants:

"we've got approximately 200 ex-pats . about 75% long- . [Assignments] . and 25% short-term . [Secondments] . One trend I have noticed is a lot of people won't be counted in that number . they'll be IT people who spend their life going from one . [county]. to the other . they work in two locations and split their time between two countries . they might go for 6 weeks here, 6 weeks there . they'll be all over the place" (HRD).

When questioned about the rationale of using expatriate, the HRD emphasised the importance of transferring knowledge over space and time: " . it is mainly people going out from London to other countries . we have all the knowledge here and we want to take it out to other places." Moreover, when questioned in more detail about the specific reasons for using expatriate staff, the importance of skills and knowledge were flagged:

"Q. So could you give me the reasons for using ex-pats?

A: One. skills. They have the relevant skills to take to another location .. that would include managerial skills as we might send someone from here to go and head up team overseas . it still has a lot to do with needing international knowledge.

Q: So Assignments are viewed positively?

A: Yes definitely . then you get 3rd country moves - for instance if someone has gone from the US to Japan and then they'll come here so they're taking knowledge from all places."

With respect to migration flows, Table 3 illustrates the net reduction of Assignments and Secondments between 1993 and 1999, but, as the HRD suggested, this could be accounted for by the exclusion of IT migration in 1999 data. As with 1993, very little data was released on destinations of migrant labour. For the investment and corporate banking arms of the group, "they are all going to Japan, Singapore, Hong Kong, New York and some European cities where we have other offices" (HRD). But, for inflows of labour, no information was given. Finally, with respect to gender, the pattern matched the 1993 survey, where "the majority of employees are male . the majority of women are not in such senior positions so you're not going to move lower grade people around" (HRD).

Westbank. In 1993, Westbank used migrant labour, which always returned to the UK, in two operational activities:

"the Expatriate move, which is usually for a period of 3 to 5 years; and the Secondment which is for shorter time periods, usually one year. The majority of Secondments are to Spain . while the majority of expatriates are located in the Channel Islands, USA and Far East." (Manager of Overseas Operations [MOO]).

Expatriates undertook "senior management positions", which usually involved "heading or deputizing in an international office" (MOO). In contrast, Secondments were sent overseas as "junior managers", with the specific remit of "attracting and developing business activities" in all facets of banking operations (MOO). All personnel were sent overseas to extend the operational capabilities of the bank, and specifically: to manage international offices on behalf of the bank; to offer a consistent approach to banking within the group; and most importantly, "to maintain and enhance the bank's profile and reputation in the locale" (MOO). Moreover, the recruitment strategy denoted the importance of skill and knowledge transfer:

"staff . [were] . selected for knowledge, organisational abilities and leadership . they will be the best person for the job. It is only high potential staff who are usually sent overseas" (MOO).

In 1993, the bank had 105 Expatriates and 10 Secondments in 14 different cities (Figure 2), with a heavy presence in private banking functions in Britain's closest offshore financial centres. All Secondments were hosted by a Spanish bank, from Madrid. With respect to gender, almost all of the expatriates were male, however, women did make up the larger proportion of Secondments. In 1993, all incoming staff were on Secondment from the Spanish bank, and specialised in information technology. As to the future, the bank was adamant that information technology would not replace the Expatriate or Secondment. The MOO stressed the importance of having migrant labour in international offices in order to disseminate the "bank's 'global' culture", with respect to organizational policies, to both local staff and clients, but conceded that "the volume of short term overseas travel had declined." Business travel had replaced very short-term postings, especially to Europe, and Secondments had become "more focussed, and undertaken to areas where business exists, rather than for previous goodwill" (MOO).

In May 1999, an interview was undertaken with the International Human Resources Director (IHRD) of the bank. The interviewee was not the same person interviewed in 1993 and much of the discussion focused upon migration flows and the role of the IHRD. In 1999, the bank still maintained Expatriates and Secondments, with the same organizational practices as reported previous. However, with respect to migration flows, data from 1999 indicates a net reduction in Expatriate postings in particular (-55), and the time-series data mapped in Figure 2, shows the withdrawal of Expatriates from several European and East Asia global cities. When pressed to account for such changes in the geography of their Expatriates, the IHRB discussed the total withdrawal of Expatriates in "retail banking", but emphasised the importance of investment, corporate and private banking Expatriate functions:

". there are a multitude of reasons for people going out. For instance in Jakarta we may need people to go and head up projects for example in some big power station, or a leverage buyout" (IHRD).

In addition, the reduction in Expatriates was attributed to the bank closing offices in Geneva and Hong Kong, and their rationalisation activities throughout East Asia in the aftermath of the financial crisis. Secondments at the Spanish bank undertook the same functions as in 1993.

Discussion

From the analysis of IPS data and TNC interviews, two major observations can be tabled which drawn together theory and practice. Regarding theory, we would like to suggest that two major threads stitch together globalization and transient professional migration, as symbiotic process, within global city IFCs.

First, professional and managerial international migration has still remained an important characteristic of the UK's economy, as we end the Twentieth Century. As we illustrated in Table 2, and Figures 1a-c, such official migration has increased during the 1990s (especially 1997), with marked magnitudes of net emigration flows of British nationals, and net immigration flows from All Commonwealth and Other Foreign countries (Figure 1e), especially males in the 24-44 age group. Moreover, research findings from TNC banks have enable us to begin to understand the specific processes, which generate professional and managerial international migration. As we have discovered in the banking longitudinal data, during the 1990s, they still employed transient professional migrants, for periods between 2-5 years on average, to embed their global culture, 'operationalise' banking functions (treasury) and represent clients, in IFCs. In essence, these banks facilitated the global 'reach' of their business through the knowledge, expertise and general 'know how' of their transient migrants, who roamed between international offices. But, the most noticeable difference in the longitudinal data was the decline in the number of Third-Country-Nationals used by the banks, and an increase in Secondments (short-term assignments). In short, the evidence presented from the interviews, provides new material to corroborate and strengthen work on globalization, TNCs, careers and transient labour (Table 1).

Second, if we examine of the geography of transient professional migration within TNC banks, we would like to suggest that it reinforces the role of such labour in the creation of cultural capital, through participation in networks. As above, an important finding was that banks used transient migration to transfer knowledge and expertise over space and time. Nothing can illustrate the importance of networking and cultural capital more clearly than the postings of transient bankers to nearby offshore financial centres (Westbank - Channel Islands) where one would expect business travel to replace such staff. Moreover, the vital roles that transient professionals have in embedding knowledge, producing knowledge and creating cultural capital from networking is epitimised by TNCs concentrating them within their the major global city IFCs (New York; Tokyo; Frankfurt; Zurich; Paris; Singapore; and Hong Kong). One would expect TNC banks not to need transient professional migrants in these locations in 1999 because these offices already have: their largest offshore employment sizes and skill bases; the highest concentrations of IT facilities; and are the most accessible offices, via air travel. Thus, apart from the agglomeration economies of these IFCs, the distribution of transient professionals in banking provides new empirical evidence to support two of the important theoretical frameworks discussed earlier in this paper. First, transient professional migrants play a vital part in the 'meeting places' and 'corporate networks' of IFCs (Amin and Thrift, 1992). And, second, transient professional migrants in banking are used by TNCs for their embedment synergies: the ability to perpetuate the organisational culture of the TNC in locations where other 'cultures' and ways of 'doing things' pre-dominate. But, more importantly, it begins to add new empirical substance to support the new economic geographers who profess about embeddedness, through their writings on networking. Such networking is the raison d'être of transient professional migrants which serves to (re)produce them anew.

Regarding method, we would like to advocate the mixed-method approach. Quite clearly, our analysis of IPS data presents an interesting, but limited, evaluation of the major trends in professional and managerial international migration. But, in order to begin to develop an understanding of 'deeper' underlying processes driving globalization and migration as symbiotic tendencies, other research strategies must be adopted. In our case, we undertook detailed semi-structured interviews with TNC banks. This had both advantages and disadvantages. With respect to the former, we were able to gain a rich and meaningful insight into transient professional migration. However, turning to the latter, problems did emerge concerning the varying quality and quantity of material collected, and negotiating access and the release of information to competitors in the public domain. The major point to make is that obtaining data from interview surveys provides an extremely rich source of material to begin to understand process: why and how 'things' are 'done.' Whilst such material can be open to problems of representativeness, reliability or incompleteness, used as a mixed-method, in this case with official IPS migration data, we were able to present a picture of both coverage (pattern) and detail (process), which could not be interpreted from solely quantitative or qualitative techniques.

Conclusions

In this paper we have attempted to increase the visibility of transient migration in the world system by presenting new research findings which investigate professional and managerial migration, within the global city IFC. We would like to make three concluding remarks. First, our study has clearly illustrated that in order to analyse in detail the processes, geographies and temporality of transient professional migration, researchers must go beyond an interpretation of official migration data. In essence, we highly recommend the use of mixed-method research in transient migration studies, which, if achieved, allows the researcher to bring together 'processes' and 'patterns'.

Second, our analysis of IPS data and the originality of the longitudinal banking interview survey has illustrated very clearly the complex relationship that exists between globalization and migration, as we enter the next century. The IPS data shows a record number of professional and managerial migration in 1997, and the 1999 survey especially points to the major organizational significance of transient professional migration. Moreover, the new and unique material collected in 1999 demonstrates the need for transient professional migration in the global service economy, despite high cost structures, and advances in information technology and business travel. In 1999, these bank still used long-term assignments and short-term secondments, as very 'visible' globalization strategies intended to extend, and maintain, core global business activities. The 1999 findings, not only strengthen our knowledge about globalization and TNCs (as discussed by John Salt and others in Table 2), but also provides new empirical evidence to corroborate the theoretical work of new economic geographers, who conceptualise embedment, knowledge and networks in IFCs. A major finding from the 1999 survey was that banks used the skills of transient professionals, usually men, in situ to create very particular cultural capital as a mechanism to extend their global reach in IFCs.

Third, and finally, looking beyond transient migration in the global city, globalization tendencies are now being identified as paramount forces in the (re)production of migration, at different scales in the world (see Table 1, and, for example, Jones and Findlay, 1998). According to Portes (1997, 1) geographies of migration can be conceptualised as "globalization from below". A term which has been coined to couch migration as a local response to the many 'top-down' tendencies of globalization in the world. Moreover, almost all the population geographers cited in this paper, are publishing rich case study material which illustrates that geographies of migration MUST be acknowledged by other academics and policy makers alike, as both integral tendencies and 'drivers' of globalization (Table 1). At present, outside of the population geography fraternity, other social scientists who profess on globalization issues are failing to acknowledge the importance of people, in their wider readings of globalization in the contemporary world.

The authors would like to thank Loughborough University and the Economic and Social Research Council's Transnational Communities Programme (L214252001), and Allan Findlay and two referees for their advice on an earlier draft of this paper.