M.Soppe et aL/Journal of Transport Geography 17(2009)10-20 11 1.1.Literature review partnerships in specific port projects,in which carriers are inter- ested in joint-investment for supporting their maritime network. Within the academic literature the topic of container port ser- Our paper demonstrates the legitimacy of such an hypothesis. vices has been considered by many studies.The involvement of showing the emerging forms(contractual and equity)of vertical shipping lines in stevedoring activities and the emergence of ded- coordination3 between carriers and stevedores and highlighting icated terminals have been discussed earlier by Slack(1993)and players which are resorting to such partnerships. Haralambides et al.(2002).Stopford (2002)showed the impor- The paper is structured as follows:Section 2 is dedicated to tance of the cost of handling operations that justifies their inter- methodological issues explaining the data used for our quantita- nalisation.Sys(2005)and Cullinane and Khanna (2000)explored tive analyses.In Section 3,we address the evolution of the con- the productivity-driven reasons related to the increasing size of tainer handling business characterised by the entry of carriers vessels.Other authors addressed the integration of terminal activ- and the extension of PTOs'networks and also the structure of the ities as a consequence of carriers'networks extension and their market comparing the demand and supply sides from a geograph- hub and spoke organization.(Baird,2006:Guy,2003;Heaver ical viewpoint.Our analyses reveal a gap in terms of geographical et al.,2000:Rimmer,2004).Analysed vertical integration strategies coverage and market maturity between carriers'demand and PTOs of carriers,highlighting the potential role of network economies in supply which is one of the elements explaining the carriers'strat- such a process.Intermodality is also among the leading factors egies vis-a-vis handling activities at the global level.Finally,Sec- pushing SLs to keep control on their terminal operations (Notte- tions 4-6 provide a deeper empirical study of the latter.aiming boom,2004a:Panayides,2002). at identifying "unrevealed"and geographically extended forms of Contrary to carriers'entry in ports,the emergence of PTOs has inter-industry co-operation between SLs and ITOs. been caught up by the mainstream literature relatively late.Early studies by Ferrari and Benacchio(2000)and Peters(2001),showed 2.Methodological notes leading players expanding in various markets.More recently,Not- teboom (2004b)discussed major PTOs in relation to the growing A quantitative analysis of carriers'traffic flow distribution over concentration in liner shipping and to the rise of global alliances. terminals is an interesting method of addressing the relationship Midoro et al.(2005).addressing the Peters'typology on terminal between terminal operators and SLs.The analysis of such figures operators'waves of internationalisation,focused on major drivers can give a novel insight into the way SLs satisfy their growing of carriers'vertical strategies in ports.Slack and Fremont (2005) needs of port services,such as the share of self-handling (through and Olivier et al.(2007)analysed the stevedoring industry clearly outlining the two dominant business models,PTO and ISL Finally WOS"and POS),the diversification of the suppliers portfolio,the degree of dependence on major third-party terminal operators Olivier (2005)approached the role of emerging (local)forms of partnerships between PTOs and ISLs and foresaw the future estab- (especially PTOs)or the relevance and the geographical spread of emerging cooperative ventures (Vs)between shipping lines and lishment of wider inter-industry relationships even at a global PTOs in ports. scale. For the purpose of this research,carriers'statistical data on traf- While much attention remains focused on global carriers and fic flows handled in each container terminal are probably the most PTOs as a customer/supplier duo or as competing port investors, relevant.Unfortunately,the latter are unavailable,being consid- no efforts have been made to evaluate the match between the de- ered as highly strategic and kept confidential by SLs.Nevertheless, mand and the supply of handling services in really quantitative terms.Moreover,the current trends reshaping the port industry the extrapolation of data on transport capacities(i.e.vessel slots) makes it possible to overcome the unavailability of carrier's (Olivier and Slack,2006)are also calling for a new understanding throughput per terminal.Fremont and Soppe(2004b)showed a of it,mostly based on terminals,instead of ports as a whole,as strong correlation (0.91)between the throughput and the trans- the new emerging "elementary unit"of analysis ("terminalisation port capacities of calling services per port.For this study we calcu- of seaports"-Slack,2007).A "terminal by terminal"approach lated such coefficient of correlation for the 145 individual facilities seems to be much more relevant to depict the existing correlations operated by ITOs.The resulting value (0.82)still shows a high de- between the top players of two industries gree of correlation and demonstrates the fairness of our approach replacing real traffic data(not available)by transport capacity. 1.2.Aim of the paper and research questions In this relation,we have collected information on maritime con- tainerised services in 2006(year-end)from the Containerisation The purpose of our research is to investigate the nature of the International on-line data set.We completed the work investigat- relationships between carriers and terminal operators in ports. ing the brand name of the different container terminal companies This paper addresses two relevant research questions.The first called at by each maritime service.This latter information is a re- one is to outline how carriers satisfy their growing need of port han- sult of an intense and deep research into all available sources dling services.We identify the ocean carriers'major handling (SLs'and terminal operators'websites,corporate interviews,Con- providers and consider whether they are predominantly interna- tainerisation International on-line,Drewry reports,professional tional or local players.We also evaluate the relevance of self-han- press releases and scientific papers).Such data on various terminal dling within the overall activity.Our methodology,which is handling suppliers,do not have a high confidentiality value at the quantitative and based on transport capacity,reveals the depen- level of a single port,but acquire a considerable value when col- dency of SLs on third-party handling providers as well as the rela- lected and aggregated on a regional or worldwide scale.In this re- tionship between carriers'maritime services and their own port network.This discussion leads to a second and even more chal- spect,accurate information on relevant shareholders has also been collected in order to weight their different financial involvement in lenging question:are carriers and stevedores really embracing a path each project. of co-operation in ports,after having fiercely battled for such a long- time?From a carrier's viewpoint,this would mean to rely on the handling capacity of the same stevedore in many ports around 3 Vertical coordination:adjustment between the different phases of the transpor- the world,thus projecting on an international scale a contractual tation chain;in this paper,such a concept is applied to the relationships between shipping lines and terminal operators in ports. relation previously managed only on a port by port basis.More- WOS:wholly owned subsidiaries. over,this much closer relationship could easily drive to equity >POS:partially owned subsidiaries.1.1. Literature review Within the academic literature the topic of container port ser￾vices has been considered by many studies. The involvement of shipping lines in stevedoring activities and the emergence of ded￾icated terminals have been discussed earlier by Slack (1993) and Haralambides et al. (2002). Stopford (2002) showed the impor￾tance of the cost of handling operations that justifies their inter￾nalisation. Sys (2005) and Cullinane and Khanna (2000) explored the productivity-driven reasons related to the increasing size of vessels. Other authors addressed the integration of terminal activ￾ities as a consequence of carriers’ networks extension and their hub and spoke organization. (Baird, 2006; Guy, 2003; Heaver et al., 2000; Rimmer, 2004). Analysed vertical integration strategies of carriers, highlighting the potential role of network economies in such a process. Intermodality is also among the leading factors pushing SLs to keep control on their terminal operations (Notte￾boom, 2004a; Panayides, 2002). Contrary to carriers’ entry in ports, the emergence of PTOs has been caught up by the mainstream literature relatively late. Early studies by Ferrari and Benacchio (2000) and Peters (2001), showed leading players expanding in various markets. More recently, Not￾teboom (2004b) discussed major PTOs in relation to the growing concentration in liner shipping and to the rise of global alliances. Midoro et al. (2005), addressing the Peters’ typology on terminal operators’ waves of internationalisation, focused on major drivers of carriers’ vertical strategies in ports. Slack and Frémont (2005) and Olivier et al. (2007) analysed the stevedoring industry clearly outlining the two dominant business models, PTO and ISL. Finally, Olivier (2005) approached the role of emerging (local) forms of partnerships between PTOs and ISLs and foresaw the future estab￾lishment of wider inter-industry relationships even at a global scale. While much attention remains focused on global carriers and PTOs as a customer/supplier duo or as competing port investors, no efforts have been made to evaluate the match between the de￾mand and the supply of handling services in really quantitative terms. Moreover, the current trends reshaping the port industry (Olivier and Slack, 2006) are also calling for a new understanding of it, mostly based on terminals, instead of ports as a whole, as the new emerging ‘‘elementary unit” of analysis (‘‘terminalisation of seaports” – Slack, 2007). A ‘‘terminal by terminal” approach seems to be much more relevant to depict the existing correlations between the top players of two industries. 1.2. Aim of the paper and research questions The purpose of our research is to investigate the nature of the relationships between carriers and terminal operators in ports. This paper addresses two relevant research questions. The first one is to outline how carriers satisfy their growing need of port han￾dling services. We identify the ocean carriers’ major handling providers and consider whether they are predominantly interna￾tional or local players. We also evaluate the relevance of self-han￾dling within the overall activity. Our methodology, which is quantitative and based on transport capacity, reveals the depen￾dency of SLs on third-party handling providers as well as the rela￾tionship between carriers’ maritime services and their own port network. This discussion leads to a second and even more chal￾lenging question: are carriers and stevedores really embracing a path of co-operation in ports, after having fiercely battled for such a long￾time? From a carrier’s viewpoint, this would mean to rely on the handling capacity of the same stevedore in many ports around the world, thus projecting on an international scale a contractual relation previously managed only on a port by port basis. More￾over, this much closer relationship could easily drive to equity partnerships in specific port projects, in which carriers are inter￾ested in joint-investment for supporting their maritime network. Our paper demonstrates the legitimacy of such an hypothesis, showing the emerging forms (contractual and equity) of vertical coordination3 between carriers and stevedores and highlighting players which are resorting to such partnerships. The paper is structured as follows: Section 2 is dedicated to methodological issues explaining the data used for our quantita￾tive analyses. In Section 3, we address the evolution of the con￾tainer handling business characterised by the entry of carriers and the extension of PTOs’ networks and also the structure of the market comparing the demand and supply sides from a geograph￾ical viewpoint. Our analyses reveal a gap in terms of geographical coverage and market maturity between carriers’ demand and PTOs’ supply which is one of the elements explaining the carriers’ strat￾egies vis-à-vis handling activities at the global level. Finally, Sec￾tions 4–6 provide a deeper empirical study of the latter, aiming at identifying ‘‘unrevealed” and geographically extended forms of inter-industry co-operation between SLs and ITOs. 2. Methodological notes A quantitative analysis of carriers’ traffic flow distribution over terminals is an interesting method of addressing the relationship between terminal operators and SLs. The analysis of such figures can give a novel insight into the way SLs satisfy their growing needs of port services, such as the share of self-handling (through WOS4 and POS5 ), the diversification of the suppliers portfolio, the degree of dependence on major third-party terminal operators (especially PTOs) or the relevance and the geographical spread of emerging cooperative ventures (JVs) between shipping lines and PTOs in ports. For the purpose of this research, carriers’ statistical data on traf- fic flows handled in each container terminal are probably the most relevant. Unfortunately, the latter are unavailable, being consid￾ered as highly strategic and kept confidential by SLs. Nevertheless, the extrapolation of data on transport capacities (i.e. vessel slots) makes it possible to overcome the unavailability of carrier’s throughput per terminal. Frémont and Soppé (2004b) showed a strong correlation (0.91) between the throughput and the trans￾port capacities of calling services per port. For this study we calcu￾lated such coefficient of correlation for the 145 individual facilities operated by ITOs. The resulting value (0.82) still shows a high de￾gree of correlation and demonstrates the fairness of our approach replacing real traffic data (not available) by transport capacity. In this relation, we have collected information on maritime con￾tainerised services in 2006 (year-end) from the Containerisation International on-line data set. We completed the work investigat￾ing the brand name of the different container terminal companies called at by each maritime service. This latter information is a re￾sult of an intense and deep research into all available sources (SLs’ and terminal operators’ websites, corporate interviews, Con￾tainerisation International on-line, Drewry reports, professional press releases and scientific papers). Such data on various terminal handling suppliers, do not have a high confidentiality value at the level of a single port, but acquire a considerable value when col￾lected and aggregated on a regional or worldwide scale. In this re￾spect, accurate information on relevant shareholders has also been collected in order to weight their different financial involvement in each project. 3 Vertical coordination: adjustment between the different phases of the transpor￾tation chain; in this paper, such a concept is applied to the relationships between shipping lines and terminal operators in ports. 4 WOS: wholly owned subsidiaries. 5 POS: partially owned subsidiaries. M. Soppé et al. / Journal of Transport Geography 17 (2009) 10–20 11