Measuring port activities and lockdown impact using automatic identification system data

Jo, Ah-Hyun; Cho, Seong-Hyun; Kim, Bo-Kyung; Kim, Kijin; Gaduena, Ammielou Working Paper Measuring port activities and lockdown impact using automatic identification system data ADB Economics Working Paper Series, No. 747 Provided in Cooperation with: Asian Development Bank (ADB), Manila Suggested Citation: Jo, Ah-Hyun; Cho, Seong-Hyun; Kim, Bo-Kyung; Kim, Kijin; Gaduena, Ammielou (2024) : Measuring port activities and lockdown impact using automatic identification system data, ADB Economics Working Paper Series, No. 747, Asian Development Bank (ADB), Manila, https://doi.org/10.22617/WPS240486-2 This Version is available at: https://hdl.handle.net/10419/310352 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/3.0/igo/ ASIAN DEVELOPMENT BANK ASIAN DEVELOPMENT BANK 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines www.adb.org MEASURING PORT ACTIVITIES AND LOCKDOWN IMPACT USING AUTOMATIC IDENTIFICATION SYSTEM DATA Ah-Hyun Jo, Seong-Hyun Cho, Bo-Kyung Kim, Kijin Kim, and Ammielou Gaduena ADB ECONOMICS WORKING PAPER SERIES NO. 747 October 2024 Measuring Port Activities and Lockdown Impact Using Automatic Identification System Data This study examines congestion in major container ports and how port characteristics and regional factors are influenced by the coronavirus disease pandemic. Port congestion indicators, such as vessel arrivals, waiting time, and service time, were developed. Descriptive analysis shows service and waiting times were significantly higher in 2021 due to global supply chain disruptions, with import/export ports more affected than transshipment hubs. Econometric analysis indicates increased mobility restrictions reduced ship calls and increased waiting times at anchorage, with significant spillover effects from neighboring ports. About the Asian Development Bank ADB is committed to achieving a prosperous, inclusive, resilient, and sustainable Asia and the Pacific, while sustaining its efforts to eradicate extreme poverty. Established in 1966, it is owned by 69 members —49 from the region. Its main instruments for helping its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and technical assistance. ASIAN DEVELOPMENT BANK The ADB Economics Working Paper Series presents research in progress to elicit comments and encourage debate on development issues in Asia and the Pacific. The views expressed are those of the authors and do not necessarily reflect the views and policies of ADB or its Board of Governors or the governments they represent. ADB Economics Working Paper Series Measuring Port Activities and Lockdown Impact Using Automatic Identification System Data Ah-Hyun Jo, Seong-Hyun Cho, Bo-Kyung Kim, Kijin Kim, and Ammielou Gaduena No. 747 | October 2024 Ah-Hyun Jo ([email protected]) is a senior researcher, Seong-Hyun Cho ([email protected]) is a researcher, and Bo-Kyung Kim ([email protected]e.kr) is an associate research fellow at the Port Research Department, Korea Maritime Institute. Kijin Kim ([email protected]) is a senior economist and Ammielou Gaduena ([email protected]) is a consultant at the Economic Research and Development Impact Department, Asian Development Bank. Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) © 2024 Asian Development Bank 6 ADB Avenue, Mandaluyong City, 1550 Metro Manila, Philippines Tel +63 2 8632 4444; Fax +63 2 8636 2444 www.adb.org Some rights reserved. Published in 2024. ISSN 2313-6537 (print), 2313-6545 (PDF) Publication Stock No. WPS240486-2 DOI: http://dx.doi.org/10.22617/WPS240486-2 The views expressed in this publication are those of the authors and do not necessarily reflect the views and policies ofthe Asian Development Bank (ADB) or its Board of Governors or the governments they represent. ADB does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use. The mention of specific companies or products of manufacturers does not imply that they are endorsed or recommended by ADB in preference to others of a similar nature that are not mentioned. By making any designation of or reference to a particular territory or geographic area inthis document, ADB does not intend to make any judgments as to the legal or other status of any territory or area. This publication is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) https://creativecommons.org/licenses/by/3.0/igo/. By using the content of this publication, you agree to be bound bytheterms of this license. For attribution, translations, adaptations, and permissions, please read the provisions andterms of use at https://www.adb.org/terms-use#openaccess. This CC license does not apply to non-ADB copyright materials in this publication. If the material is attributed toanother source, please contact the copyright owner or publisher of that source for permission to reproduce it. ADB cannot be held liable for any claims that arise as a result of your use of the material. Please contact [email protected] if you have questions or comments with respect to content, or if you wish toobtain copyright permission for your intended use that does not fall within these terms, or for permission to use theADB logo. Corrigenda to ADB publications may be found at http://www.adb.org/publications/corrigenda. Note: ADB recognizes “China” as the People’s Republic of China; “Hong Kong” as Hong Kong, China; and “Korea” as the Republic of Korea. ABSTRACT This study measures congestion in major container ports and investigates how port characteristics and regional factors influence congestion during the COVID-19 periods. We develop port congestion indicators, including vessel arrivals, vessels staying in a period, waiting time, and service time. First, descriptive analysis reveals significantly higher service and waiting times in 2021 due to global supply chain disruptions. Import/export-focused ports were more affected than transshipment hubs. Short-term events such as labor strikes also substantially impacted port congestion. Next, using econometric analyses, estimated impulse response functions indicate a decrease in the number of ship calls following increased mobility restrictions, with average waiting time at anchorage promptly increasing, while average service time at berth was comparatively less affected by the mobility measures. Additionally, we find significant mobility impacts from neighboring ports, comparable in magnitude to those from the respective ports. Keywords: container port, automatic identification system, lockdown, impulse-response JEL codes: R41, L81, F14, C32 1. Introduction The coronavirus disease (COVID-19) pandemic caused inefficiencies in ports around the world. Ports were temporarily closed or suspended due to the infection of stevedores with COVID-19, and the labor shortage at ports intensified, paralyzing port functions for a long time. The labor shortage was especially severe at ports in the United States and Europe that rely on foreign labor, as closed borders made it impossible to find those workers. Along with the disruption of port operations, the temporary surge in demand for goods due to stimulus policies in various countries led to an increase in port traffic, further exacerbating congestion and congestion at ports. Congestion in ports undermines their services, making it impossible for them to fulfill their most essential function: the normal loading and unloading of cargo. Regions such as North America and Western Europe, which are highly dependent on foreign workers, have experienced severe deterioration in port functions due to labor shortages. The inability of foreign workers to return to ports during COVID led to shortages of truck drivers and stevedores. In October 2021, the European Union was short 400,000 workers and in the second half of 2021, job openings for port workers in the United States (US) exceeded 600,000 as employers struggled to find workers. Identifying the factors that affected port congestion, especially during the COVID-19 pandemic, can provide insights into what factors are important for efficient service delivery and what measures ports should take to provide normal service. This study aims to measure port congestion for major container ports and to verify how port characteristics affect congestion. To measure port congestion the study uses Automatic Identification System (AIS) data and corresponding ship register data.1 AIS equipment on vessels records ship movement data such as location, speed, direction, and navigational status every few seconds. The main objectives of the study are as follows. First, we develop port congestion indicators, including the number of vessel arrivals, total number of vessels that stayed in a week, average waiting time, and average service time, through AIS data for the world’s top 20 container ports from 2019 to 2022. Second, through descriptive analysis, the study scrutinizes how levels changed before and during COVID-19. We also categorize them by region and major port characteristics to understand how congestion levels vary according to those factors. Finally, to capture the dynamic lockdown impact embedded in the congestion indicators, we employ econometric methods and estimate impulse-response functions using the weekly data from 2020 to 2022. This study adds value to the literature by calculating congestion levels for major ports worldwide to understand how congestion levels are changing and how they vary. This approach differs from previous studies, which primarily focused on providing information to be evaluated and monitored by port authorities and stakeholders. Also, this study demonstrates significant methodological advancements in port congestion measurement. It introduces service time as a new congestion indicator, focuses analysis on vessels actually operating within the port, and expands the scope to include unofficial anchorages. While the methodology is currently developed for container port operations, the methodologies can be expanded to passenger vessels and bulk cargo operations. We categorized distinct congestion response patterns among ports and found that these are influenced by regional factors or their cargo characteristics. For example, the indicators show that import/export-focused ports were more affected than transshipment hubs during the pandemic, implying more severe bottlenecks and disruptions at the ends of maritime supply chains. The 1 AIS was originally developed to ensure the safety and security of maritime vessel journeys. Since December 2004, the International Maritime Organization has mandated installation of AIS equipment for vessels meeting specific criteria: all ships engaged in international voyages with gross tonnage exceeding 300, cargo ships on domestic voyages with gross tonnage exceeding 500, and all passenger ships (IMO 2002). 2 econometric results suggest that stricter restrictions reduced the number of ship calls, with estimated impulse response functions showing a prompt increase in average waiting time at anchorage, while average service time at berth was less impacted by mobility measures. Additionally, the study highlights significant mobility effects from neighboring ports, similar in scale to those from the respective ports. Furthermore, ports with lower cargo volume and connectivity experienced more significant declines in ship calls, whereas waiting times increased more notably in ports with higher cargo volume and connectivity. In the paper, section 2 reviews studies using AIS data, port congestion measurement, and the pandemic's impact on port activities. Section 3 covers data and methodology, including the frameworks for indicator development and econometric modeling. Section 4 reviews port indicators and conduct descriptive analysis based on major port characteristics, along with the estimated impact of mobility restrictions on these indicators. Section 5 summarizes and details policy implications. 2. Literature Review As the accessibility of AIS data has improved in recent years, it has been increasingly utilized in various studies across the shipping, ports, logistics, and maritime fields. This surge in usage is due to advancements in technology that have enhanced collection and availability of a wide range of ship and port data, which were previously considered limited. The number of papers using AIS has increased significantly since 2012, and they cover a wide range of fields, including ship collisions, emissions estimation, ship operation monitoring, and logistics and trade analysis (Svanberg et al. 2019). Studies in the port sector have been conducted mainly on the calculation of ship air emissions in port areas and ship arrival and departure performance in ports, and identified maritime transportation connectivity (Figure 1). These port studies have been expanded into more advanced topics based on collected data rather than using AIS collected data as it is (Yang et al 2019). As AIS research develops, port congestion is one area that can be addressed. Figure 1: Scope of AIS Research Note: Divided into basic application (BA) of AIS data, extended application (EA), and advanced application (AA). The numbers in parentheses represent the number of papers published in the Journal of Navigation and Marine Pollution Bulletin from 2003 to 2008. Source: Yang et al. (2019). In the past, studies measuring congestion relied on the statistics provided by port authorities to estimate congestion using waiting time models, simulations, etc. (Yeo, Roe, and Soak 2007); Oyatoye et al. (2011). However, these studies have reliability limitations due to the use of diverse data sources, each with varying scopes and standards. This inconsistency complicates the 3 simultaneous analysis of multiple ports. Recently, methodologies for estimating congestion using AIS data have been developed. This is based on the extensive real-time location data provided by AIS, which can be matched to locations within the port to assess its level of congestion. Research on using AIS to estimate congestion has only recently emerged. Abualhaol et al. (2018) analyzed congestion levels at the Port of Singapore; Port of Hong Kong; and Port of Halifax in Canada, in 2015, and presented the congestion characteristics of specific port areas as indicators. Peng et al. (2023) analyzed about one month of AIS data in 2017 to compare the congestion levels of 20 major ports. They developed a model to predict congestion in the ports of Shanghai, Singapore, and Ningbo in the People’s Republic of China, arguing for the need to build a system for congestion monitoring using the AIS data. Chen et al. (2023) calculated congestion using AIS similarly to the above study in order to establish a system for monitoring and evaluating the congestion level of ships in a port. Examining the Los Angeles/Long Beach ports and data for one year (2021), the study presents useful implications by matching the calculated congestion level with ship operating costs to calculate the cost of congestion. A body of literature utilizing the AIS data finds that COVID-19 containment measures significantly affected global marine traffic, the predominant mode of cross-border trade. March et al. (2021) find a significant negative effect of the stringency of COVID-19-related measures to marine traffic density, and that the pandemic led to a significant and sustained slowdown in marine vessel activity along well-established maritime transport routes in Asia, Africa, and Europe. Millefiori et al. (2021) find a marked increase in the number of idle ships across all ship types and markets and a substantial decline in vessel mobility in the same period. Satellite data also showed a reduction in container ship sailings to destinations with crew-change restrictions from local COVID-19 containment policies (Heiland and Ulltveit-Moe 2020). Port performance is generally known to be significantly influenced by the quality of port infrastructure and operation efficiency. Inefficient port operations and management can increase time costs and consequently, transport costs, that are later reflected in higher freight charges, storage costs, and brokerage fees (Abe and Wilson 2009). While port performance indicators can indicate additional trade costs and barriers to trade, empirical studies are lacking that address the effects of port characteristics during the pandemic. 3. Data and Methodology 3.1 Data This study focuses on estimating port congestion, a metric intricately linked to both the number of ships present in the port area and their duration of stay. VesselsValue provides the AIS data, with records of containerships spanning 2019 to 2022. The top 20 ports are selected based on their throughput in 2022. The data contain Maritime Mobile Service Identities, call signs, International Maritime Organization numbers, latitudes, longitudes, speed over ground, course over ground, and the navigational conditions. Additionally, they are matched with ship register data by International Maritime Organization numbers, providing comprehensive ship specifications, including ship type, gross tonnage, twenty-foot equivalent unit (TEU) capacity, ship operator, and other relevant information. For the analysis, we first selected the top 30 ports by volume to analyze congestion. Of these, 19 ports were selected for congestion analysis, with characteristics of the ports such as region and country, and cargo handling such as transshipment versus import/export cargo. People’s Republic of China (PRC) ports make up the majority of the top 30 and 6 of the top 10 (Port of Hong Kong) (Table 1). 4 Table 1: List of Global Container Ports Rank ('20) Port (Sub)region Throughput (‘000 TEU) Selected for Analysis 2019 2020 2021 1 Shanghai East Asia 43,308 43,503 47,033 ○ 2 Singapore Southeast Asia 37,196 36,871 37,468 ○ 3 Ningbo East Asia 27,451 28,709 31,077 ○ 4 Shenzhen East Asia 25,548 26,548 28,770 ○ 5 Guangzhou East Asia 22,747 23,186 24,180 ○ 6 Qingdao East Asia 20,996 22,040 23,700 ○ 7 Busan East Asia 21,992 21,824 22,706 ○ 8 Tianjin East Asia 17,345 18,351 20,244 ○ 9 Hong Kong East Asia 18,303 17,969 17,798 ○ 10 Rotterdam Europe 14,811 14,349 15,300 ○ 11 Port Kelang Southeast Asia 13,581 13,244 13,724 ○ 12 Antwerp Europe 11,860 12,023 12,020 ○ 13 Xiamen East Asia 11,081 11,463 12,028 ○ 14 Dubai Middle East 14,802 11,329 13,788 ○ 15 Tanjung Pelepas Southeast Asia 9,077 9,846 11,200 ○ 16 Saigon Southeast Asia 7,554 9,724 10,385 17 Kaohsiung East Asia 10,429 9,622 9,864 ○ 18 Los Angeles North America 9,338 9,213 10,678 ○ 19 Hamburg Europe 9,258 8,522 8,708 ○ 20 Long Beach North America 7,747 7,661 9,501 ○ 21 Laem Chabang Southeast Asia 7,981 7,598 8,335 22 NY/NJ North America 7,471 7,586 8,986 ○ 23 Tanger Africa 4,802 5,771 7,174 24 Piraeus Europe 5,646 5,376 5,312 25 Valencia Europe 5,421 5,214 5,604 26 Dalian East Asia 8,760 5,110 3,670 27 Algeciras Europe 5,125 5,108 4,797 28 Bremerhaven Europe 4,857 4,771 5,019 29 Jeddah Middle East 4,434 4,737 4,739 30 Savannah North America 4,599 4,682 5,613 TEU = twenty-foot equivalent unit. Source: ISL, Shipping Statistics and Market Reviews 2020–2022 (each year publication). 3.2. Framework for Estimating Port Congestion Assessing port congestion with AIS data involves three primary steps: (i) defining the coordinates for anchorages and berths, (ii) identifying port calls, and (iii) quantifying port congestion (see Appendix 3 for the detailed algorithm). 11 Figure 7: Ratio of Total Number of Vessels Stayed to Total Number of Vessel Arrivals Source: Authors’ calculations based on AIS data. Analysis of vessel size distribution at major European and US ports reveals distinct patterns (Figure 8). Vessels were categorized as feeder (up to 3,000 TEU), intermediate (3,000–8 ,000 TEU), and Neopanamax (over 8,000 TEU). US ports (Los Angeles/Long Beach, New York/New Jersey) predominantly receive Neopanamax vessels, indicating their primary function as final destinations for large container ships carrying import/export freight, while European ports (Rotterdam, Hamburg, Antwerp) show a more balanced distribution of Feeder and Neopanamax vessels. Figure 9 shows the number of vessels stayed by size. The ports of Los Angeles/Long Beach and New York/New Jersey maintain Neopanamax dominance. Los Angeles/Long Beach experienced a notable increase in Intermediate and Neopanamax vessels stayed in 2021, indicating congestion issues affecting larger ships. European ports display more consistent patterns across all vessel sizes. 0 20 40 60 80 100 120 140 160 180 200 1.40 1.45 1.50 1.55 1.60 1.65 2019 2020 2021 2022 number of vessels (thousands) ratio (A) Total number of vessels stayed (B) Total number of vessel arrivals Ratio (A/B) 12 Figure 8: Annual Vessel Arrivals by Size, North America and European Union LALB = Los Angeles/Long Beach, NYNJ = New York/New Jersey, VAs = Vessel Arrivals. Source: Authors’ calculations based on AIS data. Figure 9: Annual Number of Vessels Stayed by Size, North America and European Union LALB = Los Angeles/Long Beach, NYNJ = New York/New Jersey, TVs = Total Number of Vessels Stayed. Source: Authors’ calculations based on AIS data. Figures 10 and 11 show vessel arrivals and stays at major Asian ports by size. Asian ports consistently receive more feeder vessels (up to 3,000 TEU) than larger ships. Compared to US and European ports, Asian ports have higher numbers of arrivals and stays across all vessel sizes. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Rotterdam Antwerp Hamburg NYNJ LALB thousands 2019 2020 2021 2022 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Rotterdam Antwerp Hamburg NYNJ LALB thousands 2019 2020 2021 2022 13 This distribution suggests Asian ports serve dual roles: as gateways for intercontinental trade and as hubs for intra-Asian cargo distribution. Figure 10: Annual Vessel Arrivals by Size, East Asia Source: Authors’ calculations based on AIS data. Figure 11: Annual Number of Vessels Stayed by Size, East Asia Source: Authors’ calculations based on AIS data. 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Shanghai Busan Singapore Hong Kong Shenzhen thousands 2019 2020 2021 2022 0.0 2.0 4.0 6.0 8.0 10.0 12.0 Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Shanghai Busan Singapore Hong Kong Shenzhen thousands Axis Title 14 4.2. Port Time of Vessel 4.2.1. Vessel service time Figure 12 compares average vessel service times across regions from 2019 to 2022. Asian ports consistently show shorter service times than non-Asian ports. Los Angeles/Long Beach had the longest average service time (80.6 hours), Hong Kong had the shortest (15.6 hours). This trend presents an interesting contrast to the rankings we observed earlier for the number of vessel arrivals and vessels stayed. European and US ports generally have fewer vessel arrivals but longer service times, whereas Asian ports have more frequent arrivals with shorter service times. This could be due to higher operational efficiency, smaller ships, or lower cargo volumes per call in Asian ports. Figure 12: Annual Vessel Average Service Time LALB = Los Angeles/Long Beach, NYNJ = New York/New Jersey. Source: Authors’ calculations based on AIS data. Figure 13 breaks down service times by vessel size for European and US ports. US ports show a clear correlation between vessel size and service time, with larger vessels requiring longer service times. Service times in Los Angeles/Long Beach increased sharply for intermediate and Neopanamax vessels in 2021–2022, highlighting congestion issues during global supply chain disruptions. European ports, however, displayed a less pronounced relationship between vessel size and service time. As shown in Figure 14, a particularly noteworthy observation is the peak in service times for Neopanamax vessels in 2021 across Asian ports. This spike aligns with the global supply chain disruptions experienced during the COVID-19 pandemic. 0 20 40 60 80 100 120 140 hours 2019 2020 2021 2022 15 Figure 13: Annual Vessel Average Service Time by Size, North America and European Union LALB = Los Angeles/Long Beach, NYNJ = New York/New Jersey. Source: Authors’ calculations based on AIS data. Figure 14: Annual Vessel Average Service Time by Size, East Asia Source: Authors’ calculations based on AIS data. Figure 15 shows Average Container Throughput per Ship (TEU/month) from 2019 to 2022 (see Appendix 6 for the details). Los Angeles/Long Beach ports demonstrate significantly higher throughput compared to other ports, averaging 9,649 TEU (LA) and 9,561 TEU (Long Beach) per vessel. This throughput steadily increased from 2019 to 2022. The high throughput at Los Angeles/Long Beach ports explains their longer average service times and fewer vessel calls relative to total cargo volume, as each vessel handles a larger quantity of containers. Figure 16 also shows a steady increase in Average Container Ship Size at Los Angeles/Long Beach ports. 0 20 40 60 80 100 120 140 160 180 Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax LALB NYNJ Antwerp Hamburg Rotterdam hours 2019 2020 2021 2022 0 5 10 15 20 25 30 35 40 Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Busan Singapore Shenzhen Shanghai Hong Kong hours 2019 2020 2021 2022 16 This trend aligns with the higher throughput per vessel and longer service times observed. The analysis of service times as a congestion indicator provides valuable insights, highlighting the importance of considering multiple, interrelated factors in assessing port performance and efficiency. Figure 15: Average Monthly Container Ship Throughput LA/LB = Los Angeles/Long Beach, NY/NJ = New York/New Jersey. Source: Korea Maritime Institute, Port Demand Analysis Center of each year. - 2 4 6 8 10 12 14 16 Jan-19 Mar-19 May-19 Jul-19 Sep-19 Nov-19 Jan-20 Mar-20 May-20 Jul-20 Sep-20 Nov-20 Jan-21 Mar-21 May-21 Jul-21 Sep-21 Nov-21 Jan-22 Mar-22 May-22 Jul-22 Sep-22 Nov-22 thousands TEU LA LB Tianjin Qingdao Dubai(Jebel Ali) NYNJ Rotterdam Shanghai Ningbo Antwerp Busan Hamburg Tanjung Pelepas Singapore Xiamen Shenzhen(Mawan) Guangzhou Port Klang Hong Kong Shenzhen(Shekou) Shenzhen(Chiwan) Kaohsiung 17 Figure 16: Average Monthly Container Ship Size LA/LB = Los Angeles/Long Beach. Source: Korea Maritime Institute, Port Demand Analysis Center of each year. On the other hand, we can expect berth productivity (container lifts per hour), a key efficiency indicator, to also affect the service time. Typically, one would expect ports with higher berth productivity to exhibit shorter average service times per vessel. From 2019 to 2022, Dubai (107 lifts/hour), Qingdao (106), Tianjin (104), and Shanghai (102) demonstrated the highest productivity, with East Asian ports generally ranking high (Appendix 7 provides details). However, average container throughput per vessel seems to have a more significant impact on service times than berth productivity. The dispersion of throughput values is much wider (highest about seven times the lowest) compared to berth productivity (highest about twice the lowest), indicating the importance of considering multiple factors when assessing port efficiency. 4.2.2. Vessel waiting time Figure 17 shows average vessel waiting times across major global ports from 2019 to 2022. Los Angeles/Long Beach had the longest average waiting time (23.3 hours), followed by Ningbo (21.6 hours), and Shanghai (18.7 hours). Los Angeles/Long Beach experienced extreme volatility, with waiting times spiking to 86.3 hours in 2021, significantly skewing its four-year average. Ningbo and Shanghai showed more consistent but elevated waiting times. Most ports experienced peak waiting times in 2021, reflecting global supply chain disruptions. Busan had the shortest average waiting time (2.6 hours), followed by Hong Kong (7.0 hours), and Hamburg (8.1 hours). Busan's low waiting time can be attributed to its unique position in AsiaEurope and Asia-North America shipping routes. As the last major port of call for many vessels departing Asia, Busan benefits from a flexible calling pattern. This strategic flexibility allows shipping lines to make last-minute decisions about calling at Busan based on current congestion levels and cargo availability, effectively minimizing waiting times. 5 6 7 8 9 10 11 thousands TEU Long Beach Los Angeles 18 Figure 17: Annual Vessel Average Waiting Time LALB = Los Angeles/Long Beach, NYNJ = New York/New Jersey. Source: Authors’ calculations based on AIS data. Figure 18 details vessel waiting times by size for European and US ports from 2019 to 2022, revealing distinct congestion patterns. Los Angeles/Long Beach port, representing the US West Coast, faced severe congestion in 2021 across all vessel sizes, with Intermediate vessels most affected. This suggests a systemic capacity issue, exacerbated by COVID-19-related import surges and supply chain disruptions. In contrast, New York/New Jersey port, representing the US East Coast, saw its peak congestion in 2022. Interestingly, the waiting times at the port in 2022 correlated clearly with vessel size, with larger vessels delayed longer. The timing of the port's congestion peak in 2022 is noteworthy and may be attributed to a shift in cargo routing strategies by logistics operators. As shippers sought to avoid the well-publicized congestion at West Coast ports, they likely diverted more cargo to East Coast ports, causing delay but significant impact on New York/New Jersey's operations. The European context, as exemplified by the Port of Hamburg, presents yet another distinct scenario. Waiting times in Hamburg increased sharply in 2022, particularly for Neopanamax vessels. This spike is likely attributable to the series of labor strikes at the port during this period. The disproportionate impact on larger vessels suggests that the strikes may have particularly affected the specialized equipment or skilled labor required to handle these ships efficiently. This situation underscores the vulnerability of port operations to labor disputes, especially when it comes to managing larger vessels. 0 10 20 30 40 50 60 70 80 90 100 hours 2019 2020 2021 2022 19 Figure 18: Annual Vessel Average Waiting Time, North America and European Union (hours) LALB = Los Angeles/Long Beach, NYNJ = New York/New Jersey. Source: Authors’ calculations based on AIS data. Figure 19 analyzes vessel waiting times by size for major Asian ports. Busan stands out with consistently low waiting times across all vessel sizes, highlighting its efficiency and strategic position. In other Asian ports, waiting times increased in 2021, though less extremely than Los Angeles/Long Beach and still lower than European ports. This indicates Asian ports' relative resilience during global supply chain disruptions. 0 10 20 30 40 50 60 70 80 90 100 Feeder intermediate neopanamax Feeder intermediate neopanamax Feeder intermediate neopanamax Feeder intermediate neopanamax Feeder intermediate neopanamax LALB Hamburg Rotterdam Antwerp NYNJ hours 2019 2020 2021 2022 20 Figure 19: Annual Vessel Average Waiting Time, Asia Source: Authors’ calculations based on AIS data. 4.3. Illustration of Congestion Analysis: Los Angeles/Long Beach and Busan Building on the congestion indicators discussed earlier, this section focuses on their development in the ports of Los Angeles/Long Beach and Busan over the sample periods. The analysis of gateway ports, such as Los Angeles/Long Beach, which serve as key entry and exit point for trade, reveals unique challenges during global supply chain disruptions. While vessel arrivals remained stable (Figure 20a), the number of vessels that stayed in port increased from mid-2020 (Figure 20b), indicating that ships were experiencing difficulties in departing the port swiftly, probably due to extended cargo handling times or prolonged waiting periods. Waiting times surged dramatically from October 2020, peaking at 150 hours in September 2021 (Figure 20c). The role of Los Angeles/Long Beach port as a primary US destination port limited cargo diversion options. The COVID-19 pandemic exacerbated congestion through labor shortages. From 2021, Los Angeles/Long Beach also saw increased average service times (Figure 20d), correlating with larger vessels and higher cargo volumes per ship. This highlights the importance of comprehensive congestion indicators. In contrast, transshipment ports like Busan showed remarkably low congestion over the study period. Busan, ranking second globally in transshipment volume, exhibited stable congestion indicators. This stability may be attributed to the inherent flexibility of transshipment operations and the competitive pressure to maintain high efficiency. Transshipment ports can more easily divert cargo during congestion. 0 5 10 15 20 25 30 35 40 Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Feeder Intermediate Neopanamax Shanghai Shenzhen Singapore Hong Kong Busan hours 2019 2020 2021 2022 27 Figure 23: Effects of a Unit Increase in the Stringency Index of Neighboring Ports (%) a. On ship calls b. On average waiting time c. On average service time 5. Conclusion This study measured port congestion in major container ports and investigated how port characteristics and regional events influence congestion, particularly during the pre-and postCOVID-19 periods. We developed port congestion indicators, including number of vessel arrivals, the total number of vessels stayed in a period, average waiting time, and average service time. Descriptive analysis of 19 global ports from 2019 to 2022 revealed several significant findings. Peak congestion was observed in 2021, coinciding with global supply chain disruptions mainly caused by the COVID-19 pandemic. Compared to US and European ports, Asian ports, on average, handle a more vessel arrivals while maintaining shorter service times. Descriptive analysis shows that gateway ports (import/export-focused) were likely to be more susceptible to congestion than transshipment hubs. For example, Los Angeles/Long Beach ports, primarily handling import/export freight, experienced severe congestion in 2021 across all vessel sizes, while congestion indicators in transshipment ports such as Busan, were stable. Labor disputes, exemplified by strikes in Hamburg, can disproportionately affect larger vessel handling. Common patterns observed through econometric methods indicate that the imposition of mobility restrictions led to a significant decline in container ship arrivals. Additionally, the analysis found a 28 statistically significant increase in average waiting times, while average service times for berthing remained largely unaffected. The results also show that the impact patterns vary based on certain port characteristics. Ship arrivals declined less in ports with higher cargo volumes and greater connectivity than in smaller, less-connected ports, while average waiting times increased more notably in these larger ports. We also find evidence of policy spillovers across ports. In response to more restrictive mobility measures imposed by neighboring ports, ship arrival numbers dropped immediately, while average waiting times at anchorage increased. These findings suggest important policy implications for port management during international emergencies. Building resilient ports requires established crisis protocols, a robust communication strategy, effective staff management, technological preparedness, and continuity in cargo flow (UNCTAD 2021). It is essential for port authorities and related stakeholders to have clear strategies in place to respond effectively and efficiently to crises, minimizing the negative impact of the shock as well as policy spillovers. Some suggested strategies are as follows: First, ports should implement robust emergency response plans that specifically address scenarios involving increased mobility restrictions. Business continuity plans and operational adjustments should ensure the continuous operation and servicing of essential cargos. Second, given the significant impact of mobility restrictions at neighboring ports, establishing communication protocols and cooperative frameworks between neighboring ports can assist in managing the flow of ships and cargo more effectively. Third, ports vulnerable to congestion should prioritize investments in infrastructure to reduce the impact. This could involve expanding berth capacities or enhancing cargo handling efficiency. Additionally, the ongoing shift toward terminal automation necessitates continuous training of port staff to improve the human-machine interface (Saanen and Shirzad 2019, Ceci 2019). Policy measures should also include support mechanisms tailored to the specific characteristics of each port. For example, smaller, less-connected ports, which are more susceptible to adverse impacts during crises, are more likely to suffer lasting effects from restrictions on transit shipping, even as hub ports begin to rebound (Tianming et al. 2021; Notteboom, Pallis, and Rodrigue 2022). 29 Appendix 1: List of Port Neighbors Port Neighbor Average Transit Days Port Neighbor Average Transit Days Antwerp Hamburg 2.1 Port Klang Shenzhen 4.8 Rotterdam 2.2 Singapore 1.9 Busan Ningbo 2.8 Tanjung Pelepas 1.6 Qingdao 2.5 Qingdao Busan 3.5 Shanghai 2.7 Singapore 14.6 Tianjin 3.2 Tianjin 2.9 Dubai Singapore 34.2 Rotterdam Antwerp 2.1 Guangzhou Hong Kong 1.4 Hamburg 2.7 Ningbo 3.1 Singapore 36.1 Tianjin 6.4 Tanjung Pelepas 22.2 Xiamen 1.9 Shanghai Busan 3.4 Hamburg Antwerp 2.3 Hong Kong 4.7 Rotterdam 2.1 Ningbo 2.1 Hong Kong Guangzhou 1.1 Qingdao 2.7 Kaohsiung 1.7 Singapore 10.6 Ningbo 3 Shenzhen Guangzhou 0.9 Singapore 5.4 Hong Kong 1.1 Xiamen 1.6 Ningbo 3 Kaohsiung Hong Kong 2.1 Singapore 4.5 Ningbo 2.8 Singapore Port Klang 2 Singapore 7 Shenzhen 4.3 Xiamen 1.6 Tanjung Pelepas 2 LALB Busan 16 Tanjung Pelepas Guangzhou 4.6 Ningbo 19.5 Singapore 2.3 NYNJ Busan 27.4 Tianjin Qingdao 6.9 Singapore 30.7 Singapore 12.9 Ningbo Busan 3.3 Xiamen Kaohsiung 1.6 Qingdao 4.8 Ningbo 2 Shanghai 2.1 Shanghai 2.9 Source: IMF PortWatch (accessed in June 2024). 30 Appendix 2: Transshipment Cargo Volume, 2021 Port Total Transshipment Transshipment Ratio Singapore 37,468 32,335 86% Busan 22,706 12,284 54% Tanjung Pelepas 11,200 10,663 95% Port Klang 13,724 8,401 61% Dubai 13,788 7,146 52% Hong Kong 17,798 6,977 39% Rotterdam 15,300 6,120 40% Shanghai 47,033 5,826 12% Ningbo 31,077 4,979 16% Kaohsiung 9,864 4,597 47% Antwerp 12,020 4,207 35% Shenzhen 28,770 3,633 13% Hamburg 8,708 3,297 38% Source: Drewry container forecaster of each year. Appendix 3: Algorithm for Calculating Port Congestion 1. Input: X: A dataframe containing AIS data with vessel positions (lon, lat), timestamps (t), speed (v), and other relevant attributes. Bport: A set of boundary coordinates defining the port′s limits. Bberth: A set of boundary coordinates defining the berth areas within the port. Banchorage: A set of boundary coordinates defining the anchorage areas within the port. 2. Data Preprocessing: 𝐅𝐅𝐅𝐅𝐅𝐅𝐭𝐭𝐅𝐅𝐅𝐅 𝐛𝐛𝐛𝐛 𝐒𝐒𝐒𝐒𝐅𝐅𝐅𝐅𝐒𝐒 X{v≤5}={(lon, lat, t, v)∈ X ∶v≤5 knots} 𝐏𝐏𝐏𝐏𝐅𝐅𝐭𝐭 𝐁𝐁𝐏𝐏𝐁𝐁𝐁𝐁𝐒𝐒𝐁𝐁𝐅𝐅𝐛𝐛 𝐄𝐄𝐄𝐄𝐭𝐭𝐅𝐅𝐁𝐁𝐄𝐄𝐭𝐭𝐅𝐅𝐏𝐏𝐁𝐁 Bport = {(lon, lat) ∈ boundary of the port} 31 𝐏𝐏𝐏𝐏𝐅𝐅𝐭𝐭 𝐖𝐖𝐅𝐅𝐒𝐒𝐅𝐅 𝐀𝐀𝐅𝐅𝐅𝐅𝐁𝐁 𝐃𝐃𝐅𝐅𝐃𝐃𝐅𝐅𝐁𝐁𝐅𝐅𝐭𝐭𝐅𝐅𝐏𝐏𝐁𝐁 Define a rectangular area Wport around the port: Wport ={(lon, lat)∶min(lon)−N ≤ lon ≤ max(lon)+ N, min(lat)− M ≤ lat ≤ max(lat)+ M} where N and M are sufficiently large numbers to cover the port area Filter X{v≤5} to obtain observations within this wide area Wport,resulting in XPortWide 𝐁𝐁𝐅𝐅𝐅𝐅𝐭𝐭𝐁𝐁 𝐁𝐁𝐁𝐁𝐒𝐒 𝐀𝐀𝐁𝐁𝐄𝐄𝐁𝐁𝐏𝐏𝐅𝐅𝐁𝐁𝐀𝐀𝐅𝐅 𝐁𝐁𝐏𝐏𝐁𝐁𝐁𝐁𝐒𝐒𝐁𝐁𝐅𝐅𝐅𝐅𝐅𝐅𝐁𝐁 Extract berth Bberth and anchorage Banchorage boundary coordinates for the port. 3. Port Call Identification: 𝐂𝐂𝐁𝐁𝐅𝐅𝐅𝐅 𝐍𝐍𝐁𝐁𝐍𝐍𝐛𝐛𝐅𝐅𝐅𝐅 𝐆𝐆𝐅𝐅𝐁𝐁𝐅𝐅𝐅𝐅𝐁𝐁𝐭𝐭𝐅𝐅𝐏𝐏𝐁𝐁 For each vessel,identify distinct port calls by calculating the time difference Δt between consecutive records: Δtj = tj+1 − tj for j = 1, 2, . . . , n −1 Assign a call number Ci to each sequence of observations with Δtj ≤ 1440 minutes. If Δtj > 1440, a new call number is assigned. 4. Berth Entry Detection: For each port call Ci,check if the vessel entered the berth area: Xberth = {(lon, lat, t) ∈ X ∶ (lon, lat) ∈ Bberth} Retain only those calls where Xberth is nonempty. 5. Time Calculations: For each port call Ci, 𝐒𝐒𝐅𝐅𝐅𝐅𝐒𝐒𝐅𝐅𝐄𝐄𝐅𝐅 𝐓𝐓𝐅𝐅𝐍𝐍𝐅𝐅 𝐓𝐓𝐁𝐁𝐅𝐅𝐅𝐅𝐒𝐒𝐅𝐅𝐄𝐄𝐅𝐅: Tservice = Σ�Δtj for (lon, lat, t)∈ Xberth� + 60 minutes 𝐖𝐖𝐁𝐁𝐅𝐅𝐭𝐭𝐅𝐅𝐁𝐁𝐀𝐀 𝐓𝐓𝐅𝐅𝐍𝐍𝐅𝐅 𝐓𝐓𝐰𝐰𝐁𝐁𝐅𝐅𝐭𝐭𝐅𝐅𝐁𝐁𝐀𝐀: Twaiting = Σ�Δtj for (lon, lat, t) ∈ Xanchorage� + 60 minutes 6. Vessel and Port Call Count Calculation: 𝐅𝐅𝐁𝐁𝐁𝐁𝐄𝐄𝐭𝐭𝐅𝐅𝐏𝐏𝐁𝐁𝐁𝐁 𝐃𝐃𝐏𝐏𝐅𝐅 𝐂𝐂𝐏𝐏𝐁𝐁𝐁𝐁𝐭𝐭𝐅𝐅𝐁𝐁𝐀𝐀 𝐔𝐔𝐁𝐁𝐅𝐅𝐔𝐔𝐁𝐁𝐅𝐅 𝐕𝐕𝐅𝐅𝐁𝐁𝐁𝐁𝐅𝐅𝐅𝐅𝐁𝐁 𝐁𝐁𝐁𝐁𝐒𝐒 𝐏𝐏𝐏𝐏𝐅𝐅𝐭𝐭 𝐂𝐂𝐁𝐁𝐅𝐅𝐅𝐅𝐁𝐁 Number of Unique Vessels: numofship(X) = |{IMOShipNok: ∀ k}| Number of Port Calls: numofcall(X) = |{Ci: ∀ i}| 32 𝐂𝐂𝐏𝐏𝐁𝐁𝐁𝐁𝐭𝐭𝐁𝐁 𝐛𝐛𝐛𝐛 𝐓𝐓𝐅𝐅𝐍𝐍𝐅𝐅 𝐏𝐏𝐅𝐅𝐅𝐅𝐅𝐅𝐏𝐏𝐒𝐒𝐁𝐁 For each time period T (e. g. , week,day,month,year) Suppose Xweek = {XT∶ T = week}, calculate: number of unique vessels 𝐁𝐁𝐁𝐁𝐍𝐍𝐏𝐏𝐃𝐃𝐁𝐁𝐁𝐁𝐅𝐅𝐒𝐒(Xweek) number of port calls 𝐁𝐁𝐁𝐁𝐍𝐍𝐏𝐏𝐃𝐃𝐄𝐄𝐁𝐁𝐅𝐅𝐅𝐅(Xweek) 7. Output: The final output includes: − a dataframe containing the calculated Tservice and Twaiting for each unique port call Ci. − Counts of unique vessels and port calls by time period T(e. g. week). IMO = International Maritime Organization. Appendix 4: Liner Shipping Connectivity Index, Fourth Quarter 2022 No. Port Index No. Port index 1 Shanghai 148 11 Kaohsiung 86 2 Ningbo 134 12 Xiamen 86 3 Singapore 128 13 Guangzhou 84 4 Busan 124 14 Hamburg 78 5 Qingdao 105 15 Dubai 78 6 Rotterdam 95 16 Tanjung Pelepas 72 7 Hong Kong 93 17 NYNJ 56 8 Port Klang 93 18 LALB 44 9 Antwerp 91 19 Tianjin 11 10 Shenzhen 87 Source: UNCTAD, LSCI: Liner shipping connectivity index (https://unctadstat.unctad.org). Appendix 5: Total Container Volume, 2021 ('000 teu/year) No. Port Volume No. Port Volume 1 Shanghai 47,033 11 Rotterdam 15,300 2 Singapore 37,468 12 Dubai 13,788 3 Ningbo 31,077 13 Port Klang 13,724 4 Shenzhen 28,770 14 Xiamen 12,028 5 Guangzhou 24,180 15 Antwerp 12,020 6 Qingdao 23,700 16 Tanjung Pelepas 11,200 7 Busan 22,706 17 Kaohsiung 9,864 8 Tianjin 20,244 18 NYNJ 8,986 9 LALB 20,179 19 Hamburg 8,708 10 Hong Kong 17,798 Source: Drewry container forecaster of each year. 33 Appendix 6: Average Container Throughput from Each Container Ship (TEU, monthly) No. Port Volume No. Port Volume 1 LA 9,649 12 Hamburg 2,997 2 LB 9,561 13 Tanjung Pelepas 2,984 3 Tianjin 4,604 14 Singapore 2,974 4 Qingdao 3,939 15 Xiamen 2,448 5 Dubai (Jebel Ali) 3,807 16 Shenzhen (Mawan) 2,390 6 NYNJ 3,790 17 Guangzhou 2,116 7 Rotterdam 3,593 18 Port Klang 1,994 8 Shanghai 3,432 19 Hong Kong 1,625 9 Ningbo 3,147 20 Shenzhen (Shekou) 1,591 10 Antwerp 3,115 21 Shenzhen (Chiwan) 1,459 11 Busan 3,014 22 Kaohsiung 1,328 Source: Korea Maritime Institute, Port Demand Analysis Center of each year. Appendix 7: Berth Productivity: Berth Moves per Hour (monthly) No. Port Volume No. Port Volume 1 Dubai (Jebel Ali) 106.82 12 Rotterdam 69.52 2 Qingdao 105.66 13 Antwerp 65.01 3 Tianjin 103.60 14 Hong Kong 64.59 4 Shanghai 101.97 15 NYNJ 64.17 5 Ningbo 85.56 16 Kaohsiung 63.70 6 Tanjung Pelepas 82.42 17 LB 61.40 7 Shenzhen (Mawan) 80.86 18 Port Klang 56.79 8 Singapore 80.28 19 LA 55.35 9 Busan 78.75 20 Hamburg 54.55 10 Guangzhou 76.90 21 Shenzhen (Chiwan) 54.49 11 Xiamen 74.78 22 Shenzhen (Shekou) 53.91 Source: Korea Maritime Institute, Port Demand Analysis Center of each year. 34 REFERENCES Abualhaol, I., R. Falcon, R. Abielmona, and E. Petriu. 2018. “Mining Port Congestion Indicators from Big AIS Data.” IEEE. https://doi.org/10.1109/IEEECONF.2018.8489187. Abe, K., and J. S. Wilson. 2009. Weathering the Storm: Investing in Port Infrastructure to Lower Trade Costs in East Asia. Policy Research Working Paper 4911. https://doi.org/10.1596/1813-9450-4911. Amador, J., C. M. Gouveia, and A. C. Pimenta. 2023. “COVID-19, Lockdowns and International Trade: Evidence from Firm-Level Data.” Empirical Economics 65: 2427–2466. https://doi.org/10.1007/s00181-023-02421-x. Ceci, Rich Sr. 2019. “Terminal Automation: The Good, The Bad and the Future.” The Journal of Ports and Terminals, edition 84. www.porttechnology.org. Chen, W., J. Chen, J. Geng, J. Ye, T. Yan, J. Shi, and J. Xu. 2023. “Monitoring and Evaluation of Ship Operation Congestion Status at Container Ports Based on AIS Data.” Ocean and Coastal Management 245: 106836. Deb, P., D. Furceri, J. Ostry, and N. Tawk. 2022. “The Economic Effects of COVID-19 Containment Measures.” Open Economies Review 33: 1–32. https://doi.org/10.1007/s11079-02109638-2. Goldstein, P., E. L. Yeyati, and L. Sartorio. 2021. “Lockdown Fatigue: The Diminishing Effects of Quarantines on the Spread of COVID-19.” Covid Economics: Vetted and Real-Time Papers 67. https://portal.cepr.org/call-papers. Google Earth. Satellite view of Port of Busan. Accessed March 2023. https://earth.google.com/web/@34.99120854,128.96730821,- 35.61808889a,110198.97827178d,35y,0h,0t,0r/data=CgRCAggBOgMKATA. Google Earth. Satellite view of Port of GuangzhouNansha Terminal. Accessed March 2023. https://earth.google.com/web/@22.66184862,113.66041827,- 2.9609263a,8764.04588952d,35y,0h,0t,0r/data=CgRCAggBOgMKATA. Google Earth. Satellite view of Port of Hamburg. Accessed March 2023. https://earth.google.com/web/@54.04884736,8.76006313,- 123.30610212a,335700.87804962d,35y,-0h,0t,0r/data=CgRCAggBOgMKATA. Google Earth. Satellite view of the North Sea coasts of Belgium and The Netherlands. Accessed March 2023. https://earth.google.com/web/@52.51359965,1.21721783,328.5214914a,754901.27913 393d,35y,0h,0t,0r/data=CgRCAggBOgMKATA. Google Earth. Satellite view of Port of Shanghai. Accessed March 2023. https://earth.google.com/web/@31.09824595,122.34060079,15.7846243a,337743.1293 85d,35y,0h,0t,0r/data=CgRCAggBOgMKATA. Heiland, I., and K. H. Ulltveit-Moe. 2020. “An Unintended Crisis in Sea Transportation due to COVID-19 Restrictions.” In COVID-19 and Trade Policy: Why Turning Inward Won’t Work, edited by R. Baldwin and S. Evenett. London: CEPR Press. https://cepr.org/chapters/unintended-crisis-sea-transportation-due-covid-19-restrictions. International Maritime Organization (IMO). 2002. Guidelines for the Onboard Operational Use of Shipborne Automatic Identification Systems (AIS). London. 35 JAG-TSC. 2023. Joint Action Group to review the impact of the COVID 19 pandemic on the world’s transport workers and the global supply chain. International Labour Organization. Jordà, Ò. 2005. “Estimation and Inference of Impulse Responses by Local Projections.” American Economic Review 95 (1): 161–182. https://doi.org/10.1257/0002828053828518. Jordà, Ò., and A. M. Taylor. 2016. “The Time for Austerity: Estimating the Average Treatment Effect of Fiscal Policy.” Economic Journal 126 (590): 219–255. https://doi.org/10.1111/ecoj.12332. Kim, B., G. Kim, and M. Kang. 2022. “Study on Comparing the Performance of Fully Automated Container Terminals during the COVID-19 Pandemic.” Sustainability 14 (15): 9415. Kim, K., J. Abesamis, and Z. Ardaniel. 2022. “Impact of COVID-19 Mobility Restrictions on Trade Facilitation at Borders in the Central Asia Regional Economic Cooperation Region.” ADB East Asia Working Paper Series No. 52. Kpodar, K., S. Quayyum, V. Gammadigbe, and M. Mlachila. 2021. Defying the Odds: Remittances During the COVID-19 Pandemic. IMF Working Papers. Washington, DC. https://www.imf.org/en/Publications/WP/Issues/2021/07/16/Defying-the-OddsRemittances-During-the-COVID-19-Pandemic-461321. Liu, X., E. Ornelas, and H. Shi. 2022. “The Trade Impact of the Covid-19 Pandemic.” The World Economy 45 (12): 3751–3779. https://doi.org/10.1111/twec.13279. March, D., K. Metcalfe, J. Tintoré, and B. Godley. 2021. “Tracking the Global Reduction of Marine Traffic during the COVID-19 Pandemic.” Nature Communications 12 (2415). https://doi.org/10.1038/s41467-021-22423-6. Millefiori, L. M., P. Braca, D. Zissis, G. Spiliopoulos, S. Marano, P. K. Willett, and S. Carniel. 2021. “COVID-19 Impact on Global Maritime Mobility.” Scientific Reports 11 (18039). Notteboom, T., A. Pallis, and J.-P. Rodrigue. 2022. Port Economics, Management and Policy. 1st ed. Routledge. https://doi.org/10.4324/9780429318184. Organisation for Economic Co-operation and Development (OECD). 2021. OECD Economic Outlook. 2021 Issue 1. Paris: OECD Publishing. https://doi.org/10.1787/edfbca02-en. Oyatoye, E. O., S. O. Adebiyi, J. C. Okoyee, and B. B. Amole. 2011. “Application of Queueing Theory to Port Congestion Problem in Nigeria.” European Journal of Business and Management 3 (8): 24–36. Peng, W., X. Bai, D. Yang, K. F. Yuen, and J. Wu. 2023. “A Deep Learning Approach for Port Congestion Estimation and Prediction.” Maritime Policy and Management 50 (7): 835–860. Saanen, Yvo, and Azadeh Shirzad. 2019. “Human-Machine Interfaces: The Key to Productivity?” The Journal of Ports and Terminals, edition 84. www.porttechnology.org. Svanberg, M., V. Santén, A. V. Hörteborn, H. A. Holm, and C. Finnsgård. 2019. “AIS in Maritime Research.” Marine Policy 106: 103520. Tianming, Gao, Vasilii Erokhin, Aleksandr Arskiy, and Mikail Khudzhatov. 2021. “Has the COVID19 Pandemic Affected Maritime Connectivity? An Estimation for China and the Polar Silk Road Countries.” Sustainability 13 (6): 3521. https://doi.org/10.3390/su13063521. United Nations Conference on Trade and Development (UNCTAD). 2021. Review of Maritime Transport 2021. Geneva. UNCTAD. 2022. COVID-19 and Maritime Transport: Navigating the Crisis and Lessons Learned. Geneva. 36 United Nations Economic Commission for Europe (UNECE). 2021. Handbook for National Master Plans for Freight Transport and Logistics. Geneva. World Trade Organization. 2023. Global Trade Outlook and Statistics. Update: October 2023. Geneva. https://www.wto.org/english/rese/bookspe/gtosupdtoct23_e.pdf. Yang, D., L. Wu, S. Wang, H. Jia, and K. X. Li. 2019. “How Big Data Enriches Maritime Research– A Critical Review of Automatic Identification System (AIS) Data Applications.” Transport Reviews 39 (6): 755–773. Yap, W. Y., and D. Yang. 2022. “Hub Port Choice and Shipping Connectivity in Southeast Asia during COVID-19 Pandemic: Implications for Post-Pandemic Competition Landscape.” Maritime Policy and Management 1–16. https://doi.org/10.1080/03088839.2022.2135179. Yeo, G. T., M. Roe, and S. M. Soak. 2007. “Evaluation of the Marine Traffic Congestion of North Harbor in Busan Port.” Journal of Waterway, Port, Coastal, and Ocean Engineering 133 (2): 87–93.