ARTICLE IN PRESS K.O'Connor /Journal of Transport Geography xxx(2009)xxx-xxx Marcotullio (2001)and "extended metropolitan area"by Ginsberg times;Pels et al.(2001)explored the outcome in a more closely etal.(1991) spaced set of airports in the San Francisco Bay Area,while Loo There are good reasons to expect that these large spatial units et al(2005)analysed these outcomes for Hong Kong and Shenzen. will play a prominent role in global logistics.One source is the con- There have been less attention to freight movements on this ceptual foundation laid by Scott who established they were inte- scale,although Schebera (2006)refers to linkages of this kind grated internally by the movement of goods(and people)and for Hong Kong-Shenzen.Hence the sea and airport infrastructure externally by world trade,both of which involve logistics activity in a global city region could be a critical factor in shaping its role That foundation is enriched by the thinking on the location of out- in global logistics activity. sourced producer services,a production arrangement central to the As noted earlier.global city regions also provide a foundation understanding of global city activity in the research of Sassen for service activity.This activity could be more mature where (1994)and extended by the research on services of Daniels and complex infrastructure provides the logistics service company Bryson(2002)and Goe et al.,(2000).Outsourcing has recently be- with an array of alternative modal choices,which it can utilize come a very significant part of logistics services as Skjoett-Larsen to meet a client's needs.Wang and Cheng(2009)have extended (2000)and Makukha and Gray(2004)have shown.Hence logistics and developed this thinking,showing that the service functions services could be a significant part of the producer service mix of of major port cities can evolve into "global supply chain manage- global city regions.Their significance might be due to the scale of ment centres"as their service activities look beyond local loading the local market in these major city regions or it might reflect spe- and unloading to include skills in finance,product planning and cial skills in managing transhipment and intermodal functions. management.That outcome could be based in part on a capacity where logistics services are in effect underpinning the hub or gate- to operate in both sea and air freight.Although there are substan- way function of the global city region. tive differences in bulk,speed and type of commodities handled There is a firm empirical base for this expected outcome too. by the two modes,it is possible that logistics service companies Hesse and Rodrigue(2004)suggest that logistics activities are lo- will have some clients who will need sea and air shipments of cated in and around the big physical nodes of seaports and airports, different goods at different times.For example,Henstra et al. but are also found at inland centres in the suburbs and in hinter- (2007)provide details of a case-study of Sony who use both sea land corridors beyond the edge of a metropolitan area.Illustrating and air transport to supply warehouses in Europe,one delivering these outcomes,Rodrigue and Notteboom(2008,p.13)have iden- base load,predicted supply,the other being used to meet unex- tified an"extended gateway"in an area of 100-150 km around the pected demand.Priemus (2001)indicated there was integration port of Antwerp,and links extending over 100 km around Rotter- between the two elements of transport infrastructure in the dam (Notteboom and Rodrigue,2008,p.65),both consistent with Netherlands.Hence logistics activity,involving both physical Rodrigue and Hess's (2007,p.116)observation that "most of the movement and value added supply chain management activity movements"to New York "involve medium distance trucking of might be better developed in a global city region than if operating a few hundred kilometers at most".The role of logistics in the glo- in a smaller city. bal city region can also be seen in the link detected between US To establish the strength of the relationship between global city west coast port traffic and the demand for warehouse space in regions and logistics activity,the research has posed three ques- west coast metropolitan areas,especially the link between the tions.First,what is the actual share of logistics activity in global ports of Los Angeles-Long Beach and the Inland Empire(Mortimer. city regions that have both sea and airports?Second,do regions 2008).That perspective is consistent with observations on the with multiple airports and seaports exert a disproportionate influ- location of physical logistics activity in big urban regions in the ence on this share?Third,to what extent do measures of physical US (King and Keating,2006),Europe (Cushman et al.,2003;Gra- logistics activity reflect measures of global city functions?The re- ham and Sahling.2004)and China (Cole et al.(2008).Hence the search reported here has developed a methodology to address location of major sea and air freight terminals within global city re- those questions. gions means there may well be"a new species of global city......a 24 hour conveyance urbanism of infrastructures,containers and 2.A methodology to link logistics and global city regions specialized vehicles...the global city as Logistics city"(Easterling. 2004.p.182). The methodology to address the issues outlined above was The issue for the research is whether this set of outcomes is developed in four stages. shaped by the location and access to large scale physical infra- structure,or whether the broader service sector in a complex ur- 2.1.Identifying global cities ban region also shapes the flows of activity.The infrastructure effect,felt via large scale sea and airport facilities,is significant The starting point here was the hierarchy of global cities devel- for global city regions as it has been shown that seaports within oped by Beaverstock et al.(2000).This is firmly rooted in a count of close proximity may collaborate in the handling of goods,and producer services in cities.There were two drawbacks to its use might in fact be served by one set of logistics service companies. One was that it was presented in categories,so it was difficult to Song (2002)suggested this outcome might have relevance in the separate out the global city status of individual cities.The second rapidly expanding Asian context where ports within a region that was that the ranking was published in 2001 and it drew upon data were once in competition might begin to co-operate.Exploration from an earlier time period.Though it might be assumed that the of his idea in the case of Shanghai and Ningbo(Wang and Olivier rank of global cities might not have changed much in this period, (2007a),Hong Kong and Shenzen(Wang and Oliver,2007b),Sin- there were some cities where logistics activity is known to be sig- gapore and Tanjung Pelapas (Tongzon,2006)and Busan and nificant,Shanghai and Dubai for example,which were likely to Gwangyang (Yeo and Cho,2007)confirms the relevance of a have become more important since 2001.The drawbacks of the large-scale urban region as a framework for logistics activity.A Beaverstock approach were overcome with the production of a similar perspective on airports was provided by de Neufville hierarchical ranking of individual cities by Mastercard Worldwide (1995).which spawned research on multiple airport regions. (2008)This was developed for a set of 50 cities in 2007,and was Fuelhart (2003,2007)has shown this effect on a small airport up-dated to 75 cities in 2008.This created a minor problem as within the catchment of larger ones in the US,noting that passen- the logistics analysis planned in the current research involved data gers can travel 70-90 miles to use different airports at different up to 2006.However,closer study of the 2008 Mastercard research Please cite this article in press as:O'Connor.K.Global city regions and the location of logistics activity.J.Transp.Geogr.(2009).doi:10.1016 jtrangeo.2009.06.015Marcotullio (2001) and ‘‘extended metropolitan area” by Ginsberg et al. (1991). There are good reasons to expect that these large spatial units will play a prominent role in global logistics. One source is the con￾ceptual foundation laid by Scott who established they were inte￾grated internally by the movement of goods (and people) and externally by world trade, both of which involve logistics activity. That foundation is enriched by the thinking on the location of out￾sourced producer services, a production arrangement central to the understanding of global city activity in the research of Sassen (1994) and extended by the research on services of Daniels and Bryson (2002) and Goe et al., (2000). Outsourcing has recently be￾come a very significant part of logistics services as Skjoett-Larsen (2000) and Makukha and Gray (2004) have shown. Hence logistics services could be a significant part of the producer service mix of global city regions. Their significance might be due to the scale of the local market in these major city regions or it might reflect spe￾cial skills in managing transhipment and intermodal functions, where logistics services are in effect underpinning the hub or gate￾way function of the global city region. There is a firm empirical base for this expected outcome too. Hesse and Rodrigue (2004) suggest that logistics activities are lo￾cated in and around the big physical nodes of seaports and airports, but are also found at inland centres in the suburbs and in hinter￾land corridors beyond the edge of a metropolitan area. Illustrating these outcomes, Rodrigue and Notteboom (2008, p. 13) have iden￾tified an ‘‘extended gateway” in an area of 100–150 km around the port of Antwerp, and links extending over 100 km around Rotter￾dam (Notteboom and Rodrigue, 2008, p. 65), both consistent with Rodrigue and Hess’s (2007, p. 116) observation that ‘‘most of the movements” to New York ‘‘involve medium distance trucking of a few hundred kilometers at most”. The role of logistics in the glo￾bal city region can also be seen in the link detected between US west coast port traffic and the demand for warehouse space in west coast metropolitan areas, especially the link between the ports of Los Angeles-Long Beach and the Inland Empire (Mortimer, 2008). That perspective is consistent with observations on the location of physical logistics activity in big urban regions in the US (King and Keating, 2006), Europe (Cushman et al., 2003; Gra￾ham and Sahling, 2004) and China (Cole et al. (2008). Hence the location of major sea and air freight terminals within global city re￾gions means there may well be ‘‘a new species of global city. . .. . ... a 24 hour conveyance urbanism of infrastructures, containers and specialized vehicles ...the global city as Logistics city” (Easterling, 2004, p. 182). The issue for the research is whether this set of outcomes is shaped by the location and access to large scale physical infra￾structure, or whether the broader service sector in a complex ur￾ban region also shapes the flows of activity. The infrastructure effect, felt via large scale sea and airport facilities, is significant for global city regions as it has been shown that seaports within close proximity may collaborate in the handling of goods, and might in fact be served by one set of logistics service companies. Song (2002) suggested this outcome might have relevance in the rapidly expanding Asian context where ports within a region that were once in competition might begin to co-operate. Exploration of his idea in the case of Shanghai and Ningbo (Wang and Olivier (2007a), Hong Kong and Shenzen (Wang and Oliver, 2007b), Sin￾gapore and Tanjung Pelapas (Tongzon, 2006) and Busan and Gwangyang (Yeo and Cho, 2007) confirms the relevance of a large-scale urban region as a framework for logistics activity. A similar perspective on airports was provided by de Neufville (1995), which spawned research on multiple airport regions. Fuelhart (2003, 2007) has shown this effect on a small airport within the catchment of larger ones in the US, noting that passen￾gers can travel 70–90 miles to use different airports at different times; Pels et al. (2001) explored the outcome in a more closely spaced set of airports in the San Francisco Bay Area, while Loo et al (2005) analysed these outcomes for Hong Kong and Shenzen. There have been less attention to freight movements on this scale, although Schebera (2006) refers to linkages of this kind for Hong Kong-Shenzen. Hence the sea and airport infrastructure in a global city region could be a critical factor in shaping its role in global logistics activity. As noted earlier, global city regions also provide a foundation for service activity. This activity could be more mature where complex infrastructure provides the logistics service company with an array of alternative modal choices, which it can utilize to meet a client’s needs. Wang and Cheng (2009) have extended and developed this thinking, showing that the service functions of major port cities can evolve into ‘‘global supply chain manage￾ment centres” as their service activities look beyond local loading and unloading to include skills in finance, product planning and management. That outcome could be based in part on a capacity to operate in both sea and air freight. Although there are substan￾tive differences in bulk, speed and type of commodities handled by the two modes, it is possible that logistics service companies will have some clients who will need sea and air shipments of different goods at different times. For example, Henstra et al. (2007) provide details of a case-study of Sony who use both sea and air transport to supply warehouses in Europe, one delivering base load, predicted supply, the other being used to meet unex￾pected demand. Priemus (2001) indicated there was integration between the two elements of transport infrastructure in the Netherlands. Hence logistics activity, involving both physical movement and value added supply chain management activity, might be better developed in a global city region than if operating in a smaller city. To establish the strength of the relationship between global city regions and logistics activity, the research has posed three ques￾tions. First, what is the actual share of logistics activity in global city regions that have both sea and airports? Second, do regions with multiple airports and seaports exert a disproportionate influ￾ence on this share? Third, to what extent do measures of physical logistics activity reflect measures of global city functions? The re￾search reported here has developed a methodology to address those questions. 2. A methodology to link logistics and global city regions The methodology to address the issues outlined above was developed in four stages. 2.1. Identifying global cities The starting point here was the hierarchy of global cities devel￾oped by Beaverstock et al. (2000). This is firmly rooted in a count of producer services in cities. There were two drawbacks to its use. One was that it was presented in categories, so it was difficult to separate out the global city status of individual cities. The second was that the ranking was published in 2001 and it drew upon data from an earlier time period. Though it might be assumed that the rank of global cities might not have changed much in this period, there were some cities where logistics activity is known to be sig￾nificant, Shanghai and Dubai for example, which were likely to have become more important since 2001. The drawbacks of the Beaverstock approach were overcome with the production of a hierarchical ranking of individual cities by Mastercard Worldwide (2008) This was developed for a set of 50 cities in 2007, and was up-dated to 75 cities in 2008. This created a minor problem as the logistics analysis planned in the current research involved data up to 2006. However, closer study of the 2008 Mastercard research 2 K. O’Connor / Journal of Transport Geography xxx (2009) xxx–xxx ARTICLE IN PRESS Please cite this article in press as: O’Connor, K. Global city regions and the location of logistics activity. J. Transp. Geogr. (2009), doi:10.1016/ j.jtrangeo.2009.06.015