The Central Asia (CA) region, comprising Afghanistan, Kazakhstan, Kyrgyz, Tajikistan, Turkmenistan and Uzbekistan, is a typical landlocked area that has suffered from limited access to global markets [1
]. CA is surrounded by Russia, China, Pakistan, Iran and the South Caucasus across the Caspian Sea. Therefore, all of the countries aim to become gateways for maritime cargo to/from the CA countries. While the current major trade partners of the CA countries are those adjacent countries, accessibility to seaports is crucial for CA countries’ sustainable economic growth. As Gallup et al. [3
] pointed out, the lower economic growth of landlocked countries can be attributed to their limited access to seaports.
Many studies have shown the importance of transport infrastructure for the international trade of CA. China’s Belt and Road Initiative (BRI), promoting regional infrastructure integration among Eurasian countries, is also concerned with transport infrastructure that would connect China to Europe. For example, regular container train services (China–Europe Railway Express) that connect Chinese cities (mainly inland cities, such as Chongqing, Chengdu, Wuhan and Xi’an) with European cities, including Russian ones, have been increasing rapidly in both their number and the amount of area that they cover in both China and Europe. The Silk Road Economic Belt concept also involves intermodal routes with land–seaport connectivity, covering Iran, the South Caucasus, Turkey and other European countries via the CA countries, without having to pass through Russia.
Meanwhile, the CA Regional Economic Cooperation (CAREC) programme, led by the Asian Development Bank (ADB), has been promoting regional cooperation and development in CA since 1997. Currently, the CAREC programme has 11 partner countries, comprising not only the CA countries, but also China (focusing on the Xinjiang Uygur Autonomous Region (XUAR)), Mongolia, Pakistan, Azerbaijan and Georgia. From the viewpoint of the international logistics environment, the XUAR is similar to the CA countries, being over 3000 km away from Chinese seaports. Thanks to the best ‘physical connectedness’ of the port-hinterland [4
] in Pakistan and China’s Belt and Road Initiative, the country’s ports, including Gwadar Port, are investing a substantial amount in their infrastructure, with the aim of becoming the new gateway seaports for CA cargo. In other words, Pakistani seaports (Karachi, Bin Qasim and Gwadar) are becoming promising gateways for CA, including the XUAR, to connect their hinterland, because most cities in CA are closer to the Pakistani seaports than the ports in Russia, China, Iran, Georgia and the Baltic countries. In fact, Pakistan joined the CAREC programme in 2010, with the aim of offering new seaport gateways for the hinterland in CA. However, the current hinterland of the Pakistani seaports extends only to Afghanistan, rather than to Pakistan itself, because of insufficient infrastructure, especially in Afghanistan, which makes it impossible to access the CA. In order for Pakistani seaports to serve the hinterland, many projects to overcome the present constraints in infrastructure in the country are either being planned or have already been implemented. For example, the China–Pakistan economic corridor (CPEC), in the context of BRI, includes launching a new seaport (Gwadar Port), enhancing domestic and international rail connections and reducing the physical and institutional barriers at the national land borders.
As ADB [5
] pointed out, most of the access routes to CA used to be for the east–west traffic, rather than the north–south traffic, even though the latter provides a shorter access to the gateway seaports. Pakistan aims to break the status quo in infrastructure investment and cross-border facilitation. While the literature, including the CPEC, has mainly highlighted the traffic analysis and role of the CPEC from the Pakistani perspective, few studies have focused on the gateway role of her seaports for CA, based on a quantitative analysis, to estimate container cargoes at major ports in Pakistan using scenario analysis. This has motivated this study to highlight a potential role of Pakistan in providing seaport gateways for CA because of her geographical advantages, which could support the sustainable development of the CA region. To the best of the authors’ knowledge, there have been no successful studies on the choice of transport routes for containers to/from CA to cover any gateway seaports across the Eurasian continent, which is due mainly to the availability of data and is partly because CA is remote in relation to world markets. To overcome the lack of data, the authors conducted a field survey for years, including several national land borders (see Table A1
and Table A2
in the Appendix A
). This study applies a two-layered network assignment model (NAM), developed from the perspective of shippers [6
], to analyse the market potential of Gwadar Port and other Pakistani ports, as gateways of the cargo to/from CA countries, including China and Russia, under six scenario conditions. The study covers the entire maritime container cargo for CA, not only China through Pakistani seaports. The authors provide a tool for quantitatively analysing them by incorporating the know-how accumulated by the analyses of different regions of the world.
The remainder of this paper is organised as follows. Section 2
details a literature review. Section 3
presents the current situation regarding the gateways available to CA among the Eurasian ports for maritime containers to/from CA. Section 4
describes the model and its input data and assesses its performance. Section 5
applies an extended two-layered network assignment model, developed from the perspective of shippers, to simulate policies related to infrastructure investment and cross-border facilitation in Pakistan and its neighbouring countries. Finally, Section 6
concludes the paper with further study issues.
2. Literature Review
Landlocked countries with a lack of transport infrastructure and poor trade facilitation face disadvantages in terms of connectivity to trade corridors and intermodal transport, causing higher transport and logistics costs [3
]. Specifically, Limao and Venables [8
] focused on the overland transport costs of landlocked countries and estimated the elasticity of their trade volume in relation to their quality of infrastructure. Arvis and Raballand [9
] highlighted the complexity of the supply chains in landlocked developing countries and clarified their logistics cost structures. Anukoonwattaka and Saggu [10
] emphasised the importance of trade policy in landlocked countries, finding that trade barriers, service trade restrictions and a poor trade facilitation performance caused high trade costs in Asian landlocked developing countries. Kashiha et al. [11
], analysing the shipment records of several countries in Europe, found that shippers in landlocked countries avoid long-haul transport, readily cross borders and place more value on transport infrastructure. Lim et al. [12
] specified key factors affecting transport corridor development in Northern Asian countries, including Mongolia, the only landlocked nation in the region. Landlocked countries recognise the importance of the geo-relationship, in terms of transport and trade corridors, between landlocked economies and their surrounding countries having seaports [13
]. Some of the aforementioned studies also indicated that landlocked countries often depend on a single transport route, which makes their economies vulnerable, emphasising the importance of transport corridor diversification in enabling them to negotiate sea access. Additionally, some papers focused on port selection and competition for cargo to/from some specific landlocked territories, such as Austria [11
], Niger [17
], Laos [18
] and the northeast part of the Southern Asian region [20
CA is a good exemplary and complicated landlocked area in the world. The World Bank [2
] categorised CA into regions far from world markets, indicating that it is comparatively difficult to reduce trade costs due to their distant hinterlands. However, CA has been recently considered as an international trade hub and as a trade partner, which is in line with the recent economic growth of its neighbours, such as China, India and Russia [21
]. In this context, some studies attempted to highlight their unsustainable trade patterns, represented by their insufficient intra-regional trade and dependence on natural resource-based exports [22
] and transport corridors [24
] in CA. Vinokurov et al. [24
] focused mainly on the access to CA from the European perspective, discussing the prioritisation of logistics investment policies, as well as the need for an integrated transport system. Kulipanova [25
] discussed the reasons for the difficulties in regional cooperation regarding transport, highlighting major physical and nonphysical barriers to international transport. Rodemann and Templar [26
] discussed the enablers and inhibitors of the promotion of rail freight transport between Europe and Asia, not only from an economic perspective, but also from political, technical, legal and environmental perspectives. Islam et al. [27
] also investigated the same issue through interview surveys with several stakeholders.
Additionally, some studies examined the bottleneck of regional trade supply chains in CA that are adjacent to emerging economies and located it at the centre of the Eurasian continent [9
]. Several studies have focused on the international logistics of a specific country or route in CA, such as Bulis and Skapars [29
], who focused on the Riga port in Latvia, as a gateway for CA cargo, and Regmi and Hanakoka [30
], who compared two specific rail routes between South Korea and European Russia. As for the quantitative approach to simulating route choice related to CA, Wang and Yeo [31
] evaluated three transport routes for second-hand cars from Korea to CA and the routes from Korea to Kazakhstan via multiple Chinese ports and land borders. Meanwhile, ADB [5
] pointed out that, given regional economic development and the accession of Turkmenistan and Pakistan to the CAREC programme, ‘an adjustment in emphasis is needed, giving as much importance to north–south and intra-regional links as with the east–west corridors between East Asia and Europe.’ In this context, Yousefi [33
] highlighted the role of Iran as a gateway for CA cargo.
There are a few studies related to Pakistan. Sayareh [34
] compared the performance of Chabahar port in Iran and that of Gwadar port in Pakistan through interviews with stakeholders. Anwar [35
] and Masood et al. [36
] discussed the potential of Pakistan as a gateway for CA. From the Chinese perspective, Khan [37
] focused on the CPEC between the XUAR and Pakistan, including a new rail connection. The CPEC is regarded as an important link and one of the key examples with which to forecast the future development of the BRI [38
]. Sheu and Kundu [41
] and Wang et al. [42
] applied the quantitative approach to simulate the choice of routes for importing oil from Middle Eastern countries into China, between the CPEC (via Pakistan), Myanmar and the direct sea route (via the Malacca Strait), although they did not consider container cargo. Shao et al. [43
] and Yang et al. [44
] focused on the rail transport that connects China with countries in the Eurasian continent and their route competitions in the BRI context, but they did not consider Pakistani seaports.
In summary, no studies have quantitatively considered CA’s connectivity to seaports in the neighbouring countries in terms of their competitiveness, focusing on Pakistani seaports, including Gwadar Port. This paper aims to fill this research gap.
3. Pakistani Seaports as a Gateway of Central Asia
The competitors of Pakistani ports, as gateways for CA cargo, are spread broadly across the Eurasian continent. Figure 1
shows the major gateway seaports and their access routes for international maritime cargo to/from CA. Broadly speaking, they have three directions in six regions, namely, (i) Far East Russia and China to the east, (ii) the Black Sea (including Georgia and Russia) and Baltic Sea (including Russia and the Baltic countries) to the west, which is a component of the Transport Corridor Europe–Caucasus–Asia (TRACECA), and (iii) Iran and Pakistan to the south. While various gateway seaports to/from CA are available across the continent, the transport time and costs associated with the three access routes are considerably high.
Among them, the Arabian Sea coast is the nearest coast for almost all CA regions, including the XUAR in China, in terms of the direct distance. However, the route to the Arabian Sea has many national borders and political/geographical difficulties, including unstable and unsafe areas, for instance, in Afghanistan and Baluchistan in Pakistan, and mountainous terrains, for instance, at the border between China and Pakistan. Guaranteeing the security of transport along this access route remains a challenge.
Currently, the sole available gateway seaport in the Arabian Sea, except for the Afghani cargo transported to the Pakistani ports, is Bandar Abbas, which is the largest seaport in Iran, located at the mouth of the Persian Gulf. It takes approximately one week to transport cargo from Tashkent to Bandar Abbas port by truck, which is almost half the time that it takes for the cargo to reach the Chinese coastal ports [46
]. It is worth noting that Chabahar port, which is located near the border with Pakistan, is expected to be a future alternative gateway among the Iranian ports.
Pakistani ports, including Karachi, Muhammad Bin Qasim and Gwadar, are becoming attractive for CA countries with respect to their potential function as new gateways, enabling the sustainable development of CA countries, since Pakistan joined the CAREC programme in 2010. However, the hinterland of these ports currently covers only Pakistan and some areas of Afghanistan because of the underdeveloped infrastructure, as well as poor security, when crossing Afghanistan, as already mentioned.
The cargo volume data and specifications of each Pakistani port are summarised in Table 1
. The development of hinterland infrastructure is planned, such as a circular railway network (Figure 2
) for Afghanistan to connect CA with the infrastructure network of Pakistan and Iran [47
Currently, approximately 60% of the total international cargo handled by the Pakistani ports is either going to or coming from the Punjab province, far from the Arabian Sea coast (over 1200 km away). However, the railway services in Pakistan are currently insufficient to meet such a transport demand. The current modal share of rail transport accounts for only 4% out of the total freight flow [50
]. This very low share of the rail transport mode is caused by (i) the lack of railroad freight cars; (ii) single tracks, even along the main railway line; and (iii) the poor service capacity in terms of port access, not to mention the substandard surface condition of the national roads.
Gwadar Port is a deep-sea port, located in the Balochistan Province of Pakistan, which is 120 km away from the border point with Iran. The initial construction was completed in 2005, and its concessional rights were transferred to China Overseas Ports Holding Company Limited in 2013 for 40 years. The port has been underused, so that it received only 145 TEUs in 2017. The first commercial container shipment was dispatched from Gwadar to UAE in March 2018 [49
]. The port development is associated directly with the CPEC [38
]. The economic corridor that will connect the port with the XUAR city of Kashi, giving China a direct access to the Arabian Sea, is planned. The corridor development covers a wide range of transport infrastructure projects, such as linking Gwadar Port with the main artery of the national highway network, developing the motorway between Gwadar city and Karachi city, expanding and reconstructing the existing main railway line in Pakistan and facilitating border-crossing transport infrastructure [51
5. Scenario Analysis for the Pakistani Intermodal Network
5.1. Scenarios Prepared
Since its accession to the CAREC programme in 2010, Pakistan has implemented several policies related to its ports becoming gateways for the sea trade of the CA countries (including the XUAR in China), such as developing Gwadar Port and investing in infrastructure along the CAREC corridors and the CPEC, as partly described in Section 3
. The scenarios to be examined in this section involve the operation of Gwadar Port and improvement of the rail connections within Pakistan and with its neighbour countries, as described in Table 3
. A base scenario (s0
) considers the original parameter settings prepared for this model—the results of which are validated in Section 4.3
. In all of the following scenarios, including scenario s1
, we assume that all regional (i.e., within the Arabian Sea, including off the coast of East Africa) LSs on the way call at Gwadar Port. Specifically, among over 1000 LSs globally, twelve services that call at any of the other Pakistani ports (i.e., Karachi and Bin Qasim), as well as any of the Persian Gulf ports (e.g., Bandar Abbas, Dubai and Khor Fakkan), are assumed to call at Gwadar Port. Additionally, the port is assumed to be connected with its hinterland by road in all of the scenarios.
Additionally, scenario s2
assume the development of the rail network in and around Pakistan. Scenario s2
assumes an improvement of the capacity in the existing rail network by increasing the service frequency, as well as a new rail construction throughout Pakistan, except for the CPEC, based on the development plan proposed by the Pakistani government (see Figure 2
). Scenario s3
additionally assumes the launch of services in the planned railway along the CAREC corridor 5, outside Pakistan (except for the CPEC), as well as those in other planned railways in Afghanistan, Iran and the South Caucasus, as shown in Figure 2
. Scenario s4
assumes the opening of the CPEC railway, in addition to scenario s3
. Two additional scenarios assume that the CAREC border barriers decline to half of the current level (s5
) and to zero (s6
), in addition to scenario s4
5.2. Port Throughputs and HT Flows Estimated
summarises the estimated annual throughputs of export and import containers for Gwadar Port in each scenario and provides breakdowns by origin/destination country in the hinterland. Table 5
shows the changes by region in the annual throughputs of export and import containers, estimated in the six scenarios, compared to those estimated in the base scenario (s0
). The results of scenario s6
are regarded as a reference, because the model computation does not converge with the given number of maximum iterations, namely, ten.
indicates that the container throughputs of Gwadar Port are estimated to exceed 10,000 TEUs for both exports and imports, if the railways are connected with the port (s2
onward). The results of scenario s2
also indicate that most containers using Gwadar Port are either going to or coming from Pakistan, while some import containers are going to Afghanistan and other CA countries (mainly to Tajikistan). Table 5
shows that some containers using Gwadar Port are shifted from other Pakistani ports, because the regional throughput for each region does not change significantly, compared to scenario s0
In contrast, in scenario s4
, assuming the launch of the CPEC railways, we find not only that the container throughputs of Gwadar Port increase by over 20,000 TEUs for both exports and imports, but also that the total throughputs of Pakistani ports increase by approximately 20,000 TEUs for both exports and imports, shifting from Chinese ports and the Black Sea ports. Particularly, the import containers that shifted to Pakistani ports in scenario s4
are mainly going to China (i.e., the XUAR) and Northern CA countries (i.e., Kazakhstan and Kyrgyz). Figure 12
shows the differences in container flow estimated in scenarios s4
, suggesting not only shifts in containers from road to rail in and around Pakistan, but also shifts from the east–west direction (to China or European Russia) to the north–south direction between Siberian Russia and Pakistani ports via Kazakhstan and China.
Additionally, the total container throughputs of Pakistan and those of Gwadar Port, in particular, increase significantly with the decline of border barriers, as shown in scenarios s5 and s6. In these scenarios, even some containers to/from Russia and the South Caucasus use Pakistani ports, in addition to the majority of containers to/from Tajikistan, Uzbekistan, Turkmenistan and Afghanistan.
5.3. Estimated MS Flows
Gwadar Port could handle over 30,000 TEUs per year (or 500 TEUs per week) in total, for both exported and imported laden containers, if the rail connectivity between Gwadar Port and each area of Pakistan was well developed and its capacity was sufficient (in scenario s2 onward). Since it is often said that 100 to 200 TEUs per week is necessary to maintain one LS to call, the above scenarios would enable several LSs (but not all LSs considered in the simulation) to sustainably call at the Gwadar Port.
shows the regional breakdown of the partner ports for maritime containers handled in three Pakistani ports in scenarios s2
. It should be noted that this is estimated based on the total throughputs, including not only exports and imports, but also transshipment containers, although they share only small portions of the total amount. It should also be noted that the partner port is defined as a port, where containers are finally discharged or first loaded into the LS network of a single company, because of the network structure of the super-network model (see Figure 3
). Specifically, the partner port is an export or import port in most cases, but some containers are occasionally transshipped there into a different company’s LS.
reveals that the regional shares of partner ports vary across Pakistani ports, while they are not significantly different among scenarios, including other scenarios that are not shown in the figure. The main partners of Gwadar Port are the Middle East and Near East, while those of Karachi port are (North- and South-) East Asia and those of Bin Qasim Port are Africa, Europe and North America. These regional differences are mainly caused by the differences in the GLS network with which each port is connected. In other words, the pattern of partner regions estimated in Gwadar Port is mainly based on the assumption that all the GLSs to call at Gwadar Port are regional services to connect with Middle East and South Asia. It should be noted that Pakistani ports (i.e., Karachi and Bin Qasim) share a significant portion of the partner ports for maritime containers handled in Gwadar Port in scenario s6
. While some containers are considered to be transshipped in these ports, other containers are coming from their hinterland, such as CA countries, and are transported to Gwadar Port by domestic shipping. This is partly because the model assumes the use of MS at least once.
In summary, the above six scenario analyses show that Gwadar Port could handle over 30,000 TEUs annually, if the rail connectivity from the port was well developed, although some cargo would shift within the Pakistani ports. If the railway is connected not only with CA, but also with China, as in scenario s4, more containers (approximately 50,000 TEUs per year for laden containers) would use Gwadar Port. These additional containers would shift partly from the gateway seaports in regions other than Pakistan. This implies that the improvement of the intermodality between ports and railways in Pakistan would only shift containers within the country, whereas Pakistan could offer gateway seaports to CA, if its international rail connections were also improved, especially those with China. Additionally, the impact on the throughputs would increase synergistically, if the border barriers decreased through an international cooperation (i.e., the CAREC programme) framework.
This study analysed the market potential of Gwadar Port and other Pakistani ports as gateways of the cargo to/from CA countries including China (the XUAR) and Russia (SFD). The results estimated with the model simulations revealed that Gwadar Port could handle a sustainable volume of containers, if the hinterland rail was well connected. If an international rail connectivity with CA via Afghanistan and China was available, Pakistani ports could play a key role as gateways for CA cargo, and the role would be intensified by lowering national border barriers under the CAREC framework. Particularly, improving the connections with China would obviously increase the cargo volume handled by Pakistani ports. In reality, the cooperation between China and Pakistan under the CPEC has been accelerated, including not only infrastructure investment in Pakistan and border facilitation between the two countries, but also the enhanced functionality of a Chinese city (Kashi), as a logistics hub. As the political instability in Afghanistan is unlikely to improve immediately, a Chinese route to connect CA with the Pakistani seaports is more realistic. Indeed, according to our interviews, some logistics companies and forwarders in CA are actively considering this option.
The major advantage of Pakistani ports is that they provide shorter distances, in terms of HT from most cities in CA, than other competing ports. If the infrastructure in and around the country was developed, and the barriers at the cross-borders were significantly reduced, the market potential of Pakistani ports, which enables the shifting of cargo from transport in the east–west directions to the north–south directions, would be realised. In addition, the three Pakistani ports, i.e., Karachi, Bin Qasim and Gwadar, have different partner regions. Particularly, Gwadar Port has a geographical advantage over the Middle East because of its MS network, although this partly depends on the GLS network with which each port is connected.
While this study showed empirical evidence regarding the market potential of Gwadar Port in terms of the container volumes drawn from the NAM, there are still many further issues to be resolved. First, a more detailed zoning system, especially for the larger countries (e.g., China and Russia) and for some European countries, is necessary to improve the model accuracy, using customs statistics and other sources in the region by country.
The second issue is the need for integrated modelling to cover both international MS and land transport cargo in the region. Herein, we considered only international maritime containers, which means that all cargo considered in the model are assumed to be moved by containerships at least once. This led to difficulties in handling the drastic change of cargo transport in this region from MS to wholly land transport by trucks or trains, and vice versa. To resolve this, the current land transport cargo demand, which is often not transported by containers, needs to be integrated in future studies.
A further issue is incorporating the networking of GLS into the NAM model. As the existing model does not consider the networking behaviour of international liner shipping companies (i.e., all GLSs are given), we should also preliminarily assume a change in the GLS network as a given input, as shown in Section 5
, for the scenario involving Gwadar Port. It is difficult to check ex post facto whether such assumptions concerning the network change are reasonable. Internalising the behaviour of the liner shipping companies is another challenging issue.