A partnership with academia

Building knowledge for trade and development

Vi Digital Library - Text Preview

Building a Dataset for Bilateral Maritime Connectivity

Study by Fugazza, Marco; Hoffmann, Jan and Razafinombana, Rado, 2013

Download original document

This paper presents a unique database reporting the shortest liner shipping routes between any pair of countries for a reference sample of 178 countries over the 2006–2012 period. Computed maritime distances are retrieved using an original database containing all existing direct liner shipping connections between pairs of countries and the corresponding sea distance. The number of transhipments necessary to connect any country pair to allow for containerizable trade is also retrieved. The contribution of this database is threefold. First, it is expected to be a useful tool for a better appreciation of transport costs and access to regular container shipping services and their impact on trade. Secondly, as presented in this paper, it helps to describe and analyse the structure of the existing global network of liner shipping services for containerizable trade, i.e. most international trade in manufactured goods. Finally, this database is expected to facilitate the construction of a bilateral liner shipping connectivity index building on UNCTAD’s original work.








POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES


STUDY SERIES No. 61






BUILDING A DATASET


FOR BILATERAL MARITIME CONNECTIVITY




by



Marco Fugazza


Jan Hoffmann


Rado Razafinombana



















UNITED NATIONS
New York and Geneva, 2013


U N I T E D N AT I O N S C O N F E R E N C E O N T R A D E A N D D E V E L O P M E N T





ii


Note


The purpose of this series of studies is to analyse policy issues and to stimulate discussions in
the area of international trade and development. The series includes studies by UNCTAD staff and by
distinguished researchers from other organizations and academia. This paper represents the personal


views of the authors only, and not the views of the UNCTAD secretariat or its member States.


This publication has not been formally edited.


The designations employed and the presentation of the material do not imply the expression of
any opinion on the part of the United Nations concerning the legal status of any country, territory, city
or area, or of authorities or concerning the delimitation of its frontiers or boundaries.



Material in this publication may be freely quoted or reprinted, but acknowledgement is
requested, together with a copy of the publication containing the quotation or reprint to be sent to the


UNCTAD secretariat at the following address:


Marco Fugazza


Trade Analysis Branch
Division on International Trade in Goods and Services, and Commodities


United Nations Conference on Trade and Development
Palais des Nations, CH-1211 Geneva 10, Switzerland


Tel: +41 22 917 5772; Fax: +41 22 917 0044
E-mail: marco.fugazza@unctad.org







Series Editor:


Victor Ognivtsev
Officer-in-Charge


Trade Analysis Branch


DITC/UNCTAD













UNCTAD/ITCD/TAB/63









UNITED NATIONS PUBLICATION


ISSN 1607-8291









© Copyright United Nations 2013
All rights reserved





iii


Abstract



This paper presents a unique database reporting the shortest liner shipping routes between


any pair of countries for a reference sample of 178 countries over the 2006–2012 period. Computed


maritime distances are retrieved using an original database containing all existing direct liner shipping


connections between pairs of countries and the corresponding sea distance. The number of


transhipments necessary to connect any country pair to allow for containerizable trade is also


retrieved. The contribution of this database is threefold. First, it is expected to be a useful tool for a


better appreciation of transport costs and access to regular container shipping services and their


impact on trade. Secondly, as presented in this paper, it helps to describe and analyse the structure of


the existing global network of liner shipping services for containerizable trade, i.e. most international


trade in manufactured goods. Finally, our database is expected to facilitate the construction of a


bilateral liner shipping connectivity index building on UNCTAD’s original work.







Keywords: Maritime Transport, Sea Distance, Containerizable Trade, Trade Costs




JEL Classification: C61, F1, L91

















iv






Acknowledgements





We are grateful to Bismark Sitorus and Jan-Willem van Hoogenhuizen for detailed feedback


and comments.

Any mistakes or errors remain the author's own.






v


Contents




1 INTRODUCTION .......................................................................................................................... 1




2 DATA AND ALGORITHM ............................................................................................................ 2




3 DESCRIPTIVE STATISTICS AND STYLIZED FACTS ................................................................ 4




3.1 Connectivity: Number of transhipments ............................................................................ 4


3.2 Sea and maritime distances .............................................................................................. 8


3.3 Trade, maritime distance and transhipments .................................................................. 12




4 APPLICATIONS AND FUTURE RESEARCH ............................................................................ 15




REFERENCES ......................................................................................................................................... 17






vi


List of figures




Graph 1. Shortest path in graph theory ................................................................................................ 4




Figure 1. Number of transhipments by country/country groups .......................................................... 6


Figure 2. Direct sea distance and maritime distance (estimated) with transhipments ....................... 11


Figure 3. Direct sea distance and maritime distance (estimated) ....................................................... 11


Figure 4. Maritime distance (estimated) and number of transhipments (country averages) .............. 12


Figure 5. Containerizable exports and maritime distance (estimated)


for liner shipping connections ............................................................................................. 13


Figure 6. Containerizable exports and number of transhipments ...................................................... 13


Figure 7. Containerizable exports and direct connections ................................................................. 14


Figure 8. Zero trade and maritime connectivity .................................................................................. 14


Figure 9. Trade imbalances and connectivity ..................................................................................... 15






List of tables




Table 1. Number of transhipments (share in total number of bilateral relationships) .......................... 5


Table 2. Top and bottom 15 countries: Average number of transhipments ....................................... 5


Table 3. Top 10 connected countries: Number of direct connections (selected years) ...................... 7


Table 4. Bottom 10 connected countries: Number of direct connections (selected years) ................ 7


Table 5. Top 10 connected countries: Number of connections


with a maximum of two transhipments (selected years) ....................................................... 8


Table 6. Bottom 10 connected countries: Number of connections


with a maximum of two transhipments (selected years): ...................................................... 8


Table 7. Maritime distance (estimated): 2006 ..................................................................................... 9


Table 8. Maritime distance (estimated): 2012 ..................................................................................... 9


Table 9. Variations in estimated maritime distances and number of transhipments ........................ 10
















Building a Dataset for Bilateral Maritime Connectivity 1


1. INTRODUCTION


Maritime transport is at the core of international trade in merchandises. Around 80 per


cent of volume of goods exchanged in the world are transported via sea (UNCTAD, 2008). Between


1970 and 2010, developing countries´ share in the volume of seaborne exports rose from just 18 per


cent to 56 per cent of the world´s total (UNCTAD, 2013).




Containerizable transport services are key for trade in manufactured goods and global value


chains. Without access to regular liner shipping services that make use of standardized sea containers,


countries cannot competitively participate in globalized production. A recent empirical study confirmed


the “[e]ffects of the Container Revolution on World Trade” (Bernhofen et al., 2013). As pointed out by


The Economist (2013), “[c]ontainers have been more important for globalization than freer trade”.




Recent literature has emphasized the importance of transport costs and infrastructure in


explaining trade and access to international markets. Different empirical strategies have been used to


produce estimates of the overall level of transport costs.




Some studies used the ratio between imports CIF and imports FOB to proxy transportation


costs, the so-called CIF–FOB ratio (e.g. Baier and Bergstrand, 2001; Hummels and Lugovskyy, 2006).


Estimates vary essentially with the level of product aggregation. A reasonable average estimate of such


ratio computed based on total imports CIF and FOB at the country level ranges between 6 per cent


and 12 per cent. At more disaggregated product levels their dispersion increases. Approximations of


CIF–FOB ratios are higher for developing than for developed regions. UNCTAD estimates that in the


last decade, freight costs amounted 6.4 per cent for developed countries’ imports as compared to 10.6


per cent for Africa (UNCTAD, 2011).




Based on the estimation of a gravity model using United States data, Anderson and Van


Wincoop (2004) found that transport costs correspond to an average ad valorem tax equivalent of 21


per cent. These 21 per cent include both directly measured freight costs and a 9 per cent tax


equivalent of the time value of goods in transit. Using a similar empirical approach, Clark et al. (2004)


reckon that for most Latin American countries, transport costs are a greater barrier to United States


markets than import tariffs. They also find that ports efficiency is an important determinant of shipping


costs.




The recent work of Arvis et al. (2013) is an extension of the contribution of Jacks et al. (2011).


As such, it represents the most comprehensive country-level analysis of trade costs and their


components to date. Their database includes 178 countries and covers the 1995–2010 period.


Estimates of trade costs are inferred from the observed pattern of production and trade across


countries. Results indicate that maritime transport connectivity and logistics performance are very


important determinants of bilateral trade costs: UNCTAD’s Liner Shipping Connectivity Index (LSCI)


and the World Bank’s Logistics Performance Index (LPI)
1
are together a more important source of


variation in trade costs than geographical distance, and the effect is particularly strong for trade


relations involving the South.




In order to facilitate further and more extensive analysis of container transport services, trade


costs and flows, we construct a unique database reporting the shortest maritime liner shipping routes


between any pair of countries for a reference sample of 178 countries over the 2006–2012 period. In


non-technical terms, a “liner shipping” service can be compared to a regular bus service, with a bus


“line”, with fixed departure times and with many other passengers on the same bus. This is comparable




1 The World Bank's Logistics Performance Index (LPI) and UNCTAD's Liner Shipping Connectivity Index (LSCI) both aim in
different ways to provide information about countries' trade competitiveness in the area of transport and logistics.
However, the scope of the activities and countries covered, as well as the measurement approach, are rather different. In
spite of these differences, both indexes are statistically positively correlated, with a partial correlation coefficient of +0.71.
Information concerning UNCTAD's LSCI is available in UNCTAD's Review of Maritime Transport. A detailed description
and data of the World Bank, LPI is available via the website http://www.worldbank.org/lpi.





2 POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES


to the liner shipping service, where your container will be on the same ship as other containers


belonging to many different owners. When we talk about liner shipping services (and the corresponding


routes and distances), we look at a network of regular container shipping services. Thanks to


containerization and the global liner shipping network, small and large importers and exporters of


finished and intermediate containerizable goods from far away countries can trade with each other,


even if their individual trade transaction would not economically justify chartering a ship to transport a


few containers from A to B. Thanks to regular container shipping services and transhipment operations


in so-called hub ports, basically all countries are today connected to each other. To illustrate the point,


think of the Paris Metro, which is also a network of “lines”, and you can calculate how many


“transhipments” you may need to get from Gare Montparnasse to Rue de la Pompe, and you can


calculate the “shortest route” to get from Gare Montparnasse to Rue de la Pompe, even if there is no


direct metro service between the two (Hoffmann, 2012).




Shortest routes are obtained by solving for the shortest path problem in the frame of the


Graph mathematical theory applying Dijkstra’s algorithm. Computed maritime liner service distances


are retrieved using an original database containing all existing direct liner service connections between


pairs of countries and the corresponding sea distance between the two countries’ respective main


container ports. If a connection is considered “direct”, it implies that there is no need for transhipment


in a third country. Sea distance between pairs of countries represents the distance separating each


coastal country’s main port(s). In the cases of some large countries with several coast lines (e.g. the


United States of America, Canada and others), the main port retained varies according to the trade


partner considered.




Our results provide some interesting insights into the structure of the global liner shipping


network. For instance, if we consider the data for 2012, about 13.3 per cent of the country pairs in our


sample are connected directly, 9.6 per cent require one transhipment, 46.4 per cent, two


transhipments and 21 per cent, three transhipments. This is to say that almost 70 per cent of country


pairs are connected with no more than two transhipments and more than 90 per cent with no more


than three transhipments.
2




The rest of the paper is organized as follows. The next section presents our basic data and the


algorithm used to compute maritime distances for connections without a direct service. Section 3


reviews some descriptive statistics and presents some stylized facts. The last section discusses


immediate applications of our dataset and possible directions for further research.






2. DATA AND ALGORITHM


The resulting dataset includes 178 countries, 33 of which are landlocked. While landlocked


countries have by definition no direct access to liner shipping services (their country level LSCI is not


computed), they do of course also trade with overseas trading partners, making use of their


neighbouring countries’ seaports. In order to be able to include landlocked countries in the analysis of


trade and trade costs, they are also included in the database on maritime distances, assigning the


distances from/to container ports in the transit country through which the largest share of overseas


trade passes.




Six years are informed over the 2006–2012 period. The year 2007 is missing. Information on


the number of transhipments necessary to connect any pair of countries is symmetric: if two


transhipments are necessary to move containers from country C to country D, then the same number


of transhipments is necessary to move containers in the opposite direction from D to C.






2 These percentages are slightly different from earlier analysis (UNCTAD, 2013) because in this paper our database
includes landlocked countries, which are connected to the global shipping network through their neighbouring transit
countries.







Building a Dataset for Bilateral Maritime Connectivity 3


The original dataset


The original dataset includes two variables for each pair of country. The first variable is the


maritime distance between the main container ports. The second variable is a dummy variable that


assumes the value 1 if a direct service between the two countries exists, and 0 otherwise. Note that


“direct” implies that there is no need for transhipment; however, the ship will usually call at other ports


en route. The information on the existence or not of a direct connection is retrieved from the UNCTAD's


Liner Shipping Connectivity Matrix (LSCM). The information contained in the latter database is obtained


annually, in the month of May, through Lloyds List Intelligence.3 The data covers the reported


deployment of all containerships at a given point in time. This methodology allows for comparisons


over time, as the sample is always complete. UNCTAD began the systematic annual gathering of data


in 2004 at the country level, and in 2006 at the pair-of-country level.




The algorithm


The original dataset informs exclusively on the existence or not of a direct connection between


two countries. This is already an important indication of a country's connectivity. However, this would


restrict the number of assessable trade relationships to 13.3 per cent of all potential trade relationships.


In order to complement the original information set we apply Dijkstra’s (1959) algorithm to our original


data. It is the most celebrated algorithm for the solution of the shortest path problem in graph theory.


For a given source (node) in a graph such as graph 1, the algorithm finds the shortest path between


that node and every other node. For example, if the nodes of the graph represent countries and edge


path costs represent sea distances between pairs of countries connected directly, Dijkstra's algorithm


can be used to find the shortest route between one country and any other country. In other words,


Dijkstra's algorithm allows us to identify the shortest route in terms of sea distance to cover


connections between any two countries. Note that the shortest route will by default be a direct


connection if it exists. Consequently, the number of transhipments necessary to connect two countries


is minimized. Graph 1 illustrates the solution for connecting country A to country F. The shortest path


goes through country D and the total sea distance covered equals 10. The total sea distance would


correspond to our measure of maritime distance. Graph 1 also illustrates the solution of the shortest


path between country E and country F. Despite the fact that total sea distance between E and F going


through countries G and D (i.e. 4+5+3) would be shorter that the direct distance between E and F (i.e.


13) the direct connection is retained by the algorithm. This hierarchy imposed to the algorithm reflects


the fact that the cost of transhipment is likely to be much larger than the cost induced by the coverage


of a longer distance but without transhipment. This constraint is in line with existing empirical findings.


The analysis of Wilmsmeier and Hoffmann (2008) suggests that transhipment has the equivalent impact


on freight rates as an increase in distance between two countries of 2,612 km.








3 Detailed information and access conditions are available through the website
http://www.lloydslist.com/ll/sector/containers/. Until 2011 the data was obtained annually in the month of July through
Containerization International On-line, which has since been incorporated into Lloyds List Intelligence.





4 POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES


Graph 1: Shortest path in graph theory


Country C


Country G


Country A


Country E


Country F


Country D


Country B5


6


7


3


A-F Shortest Path


13


5


4


E-F Shortest Path






3. DESCRIPTIVE STATISTICS AND STYLIZED FACTS


This section presents and briefly discusses some descriptive statistics and possible stylized


facts using data on computed maritime distances and nature of connections. As mentioned before,


178 countries make our reference sample. Information is available for the year 2006, and for the years


from 2008 to 2012.




3.1 CONNECTIVITY: NUMBER OF TRANSHIPMENTS


Table 1 characterizes the nature of the connection between pairs of countries across years.


Figures correspond to the share of the number of transhipments necessary to connect two countries in


the overall number of country–pairs connections present in the sample, that is 178,177 ( = 31,506) each


year.




Over the whole period, on average about 13 per cent of country pairs are connected directly,


about 10 per cent need one transhipment, about 49 per cent, two transhipments and about 21 per


cent, three transhipments. This is to say that about 72 per cent of country pairs are connected with no


more than two transhipments and around 93 per cent ,with no more than three transhipments.










Building a Dataset for Bilateral Maritime Connectivity 5


Table 1: Number of transhipments


(Share in total number of bilateral relationships)


2006 2008 2009 2010 2011 2012


0 13.3 13.8 13.2 13.6 13.3 13.3


1 9.5 9.9 9.7 10.3 9.7 9.6


2 49.0 49.6 49.5 50.0 49.0 46.4


3 21.2 22.0 21.6 20.2 20.8 21.0


4 5.7 4.4 5.2 5.2 6.5 6.9


5 1.0 0.3 0.6 0.6 0.8 1.9


6 0.3 0.0 0.2 0.1 0.0 0.6


7 0.1 0.0 0.0 0.0 0.0 0.3


8 0.0 0.0 0.0 0.0 0.0 0.1


Total 100.0 100.0 100.0 100.0 100.0 100.0




Looking at the average number of connections at the country level over the whole period of


time as reported in Table 2 (left quadrant) we observe that this characteristic is actually common to


several large advanced economies. Indeed, the United Kingdom of Great Britain and Northern Ireland


is the country with the smallest average number of transhipments, followed by France, Belgium,


Germany and three other countries of the European Union. This ranking could be the result of a strong


intra-European Union trade effect. Nevertheless even when trade relationships with other members of


the Union are not included, those European countries stay among the top 10 country list. The other top


15 countries are the United States and seven East Asian countries. There is again a clear intraregional


effect within the latter group of countries.




The right quadrant of Table 2 contains the corresponding bottom 15 countries. The


geographical composition is more heterogeneous and all continents are represented. The bottom list is


not only made of landlocked countries and small island States.






Table 2: Top and bottom 15 countries: Average number of transhipments


Top 15 Mean Bottom 15 Mean


GBR 0.73 RWA 3.15


FRA 0.79 MWI 3.15


BEL 0.84 ZMB 3.15


DEU 0.87 BOL 3.16


NLD 0.88 ISL 3.16


ITA 0.92 TKM 3.20


ESP 0.93 NER 3.20


CHN, HKG SAR 0.95 BLZ 3.23


CHN 0.97 SVK 3.31


USA 0.98 HUN 3.31


KOR 1.07 BLR 3.32


MYS 1.11 NRU 3.42


SGP 1.13 MLI 3.53


CHN, TWN Province of 1.19 MDA 3.62


JPN 1.29 ARM 4.10


Note: Names of countries, territories or areas of geographical interest and their subdivisions are listed here
according to the country alpha-3-codes established by the International Organization for Standardization
(ISO).





6 POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES


Figure 1: Number of transhipments by country/country groups


0
.


2
.


4
.


6
Sh


ar
e


2006 2008 2010 2012


0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7


Africa
0


.
2


.
4


.
6


Sh
ar


e


2006 2008 2010 2012


0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7


America


0.
1.2


.
3.4


.
5


Sh
ar


e


2006 2008 2010 2012


0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7


Asia


Number of transhipments


0
.


1
.


2
.


3
.


4
.


5
Sh


ar
e


2006 2008 2010 2012


0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7


Pacific


0
.


1
.


2
.


3
.


4
.


5
Sh


ar
e


2006 2008 2010 2012


0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7


Europe


Number of transhipments


0.
1.2


.
3.4


.
5


Sh
ar


e


2006 2008 2010 2012


0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4


USA


0
.


2
.


4
.


6
Sh


ar
e


2006 2008 2010 2012


0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4


CAN


0.
1.


2.
3.


4
Sh


ar
e


2006 2008 2010 2012


0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7


EUR


Number of transhipments


0.
1.2


.
3.4


.
5


Sh
ar


e


2006 2008 2010 2012


0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4


JPN


0
.


2.
4.


6.
8


Sh
ar


e


2006 2008 2010 2012


0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7


AUS


0.
2.


4.
6.


8
Sh


ar
e


2006 2008 2010 2012


0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7


NZL


Number of transhipments


0
.


1
.


2
.


3
.


4
.


5
Sh


ar
e


2006 2008 2010 2012


0 1 2 3 4 0 1 2 3 4 0 1 2 3 4 0 1 2 3 4


CHN


0
.


1
.


2
.


3
.


4
.


5
Sh


ar
e


2006 2008 2010 2012


0 1 2 3 0 1 2 3 0 1 2 3 0 1 2 3


IND


Number of transhipments




Note: Names of countries, territories or areas of geographical interest and their subdivisions are listed here
according to ISO country alpha-3-codes.







Building a Dataset for Bilateral Maritime Connectivity 7


Table 3 and Table 4 report the top and bottom 10 countries respectively in terms of number of


direct connections. As far as the top countries are concerned figures do confirm what was shown in


the previous table. Great Britain enjoys the largest number of direct connections in all four years


reported despite the fact that between 2006 and 2012 it has lost 10 per cent of them. No general trend


pops up. Some countries have seen the number of direct connections increasing others have seen it


decreasing (e.g. Great Britain). The group composition has only marginally changed over the period


with the exit of Italy on one hand and the entry of Malaysia on the other end. This is somehow in


contrast with the bottom 10 country group. Only five countries stayed in the latter group over the whole


period.




Table 3: Top 10 connected countries: Number of direct connections (selected years)


2006 2008 2010 2012


GBR 105 GBR 108 GBR 99 GBR 93


BEL 98 FRA 99 FRA 96 FRA 92


FRA 96 BEL 97 BEL 92 USA 91


DEU 93 DEU 96 CHN, HKG SAR 89 NLD 88


USA 90 ESP 91 CHN 88 BEL 88


ESP 89 ITA 90 USA 86 CHN 86


NLD 89 USA 89 NLD 86 CHN, HKG SAR 85


ITA 84 NLD 87 DEU 85 ESP 83


CHN, HKG SAR 82 CHN 81 ITA 79 MYS 82


CHN 77 CHN, HKG SAR 81 ESP 79 DEU 81


Note: Names of countries, territories or areas of geographical interest and their subdivisions are listed here
according to ISO country alpha-3-codes.



Table 4: Bottom 10 connected countries: Number of direct connections (selected years)


2006 2008 2010 2012


NRU 1 NRU 1 ALB 1 ALB 1


ALB 1 IRQ 2 MMR 2 QAT 2


MMR 2 QAT 2 IRQ 3 MMR 2


BHR 3 PLW 3 QAT 3 IRQ 3


IRQ 4 SOM 3 NRU 3 BRN 3


QAT 4 BHR 3 MDV 4 NRU 3


PLW 4 ALB 3 BGD 5 BGD 4


BLZ 4 KWT 4 PLW 5 MDV 4


BRN 4 SYC 4 SOM 6 PLW 5


KWT 4 BGD 4 BRN 6 SOM 6


Note: Names of countries, territories or areas of geographical interest and their subdivisions are listed here
according to ISO country alpha-3-codes.




A noticeable fact is the significant decrease after 2008 in the number of direct connections


enjoyed by the group of the top 10. This could be clearly seen as a consequence of the collapse of


world demand in the aftermath of the financial crisis started at the end of the year 2007. Counting the


number of connections with a maximum of two transhipments generates slightly different results at


both the top and the bottom of the country ranking. As shown in table 5, economies such as


Singapore, Brazil, Egypt, Taiwan Province of China and Portugal, appear at least once among the list of


the top 10. The composition of the worst performer country group varies quite significantly over the


period, as shown in Table 6. In addition, many of these countries were not in the bottom group when


considering the number of direct connections. The maximum number of connexions is observed for the


United Kingdom in 2006 and equals 177. The lowest number of connections is observed for Nauru in


2010 and equals 29.




In general, allowing for two transhipments considerably increases the number of reachable


destinations especially for the most remote economies such as Albania and Nauru. In the former case,


this is explained by the proximity of an extremely well-connected country such as Italy, which acts as a





8 POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES


transit export platform. Nauru, despite an exponential increase of potential connections, remains the


most remote economy.




Table 5: Top 10 connected countries: Number of connections with a maximum of two


transhipments (selected years)


2006 2008 2010 2012
GBR 177 ESP 176 ESP 176 GBR 174


ESP 176 GBR 176 GBR 176 NLD 174


NLD 176 NLD 176 NLD 176 CHN, TWN Province of 173


ITA 174 BEL 175 BEL 175 MYS 171


BEL 173 FRA 174 PRT 174 KOR 171


FRA 173 ITA 173 FRA 174 FRA 171


CHN, TWN Province of 172 DEU 171 BRA 174 ESP 171


DEU 171 CHN, TWN Province of 170 KOR 173 CHN, HKG SAR 171


CHN, HKG SAR 169 PRT 170 CHN, HKG SAR 173 BEL 169


SGP 168 CHN, HKG SAR 170 EGY 173 DEU 168


Note: Names of countries, territories or areas of geographical interest and their subdivisions are listed here
according to ISO country alpha-3-codes.






Table 6: Bottom 10 connected countries: Number of connections with a maximum of two


transhipments (selected years)


2006 2008 2010 2012


BLZ 34 NRU 41 NRU 29 NRU 30


NRU 36 BLZ 41 ARM 34 LTU 32


COD 38 ISL 41 IRQ 39 ISL 33


LVA 49 IRQ 44 GEO 42 EST 33


ISL 49 LVA 45 LTU 45 LVA 36


SUR 50 SUR 49 LVA 46 SLV 36


SOM 51 GUY 49 EST 47 ARM 36


ARM 54 SYC 52 ISL 47 NIC 38


MDV 57 SOM 54 PLW 48 ABW 39


GUY 59 HTI 56 BLZ 51 PLW 42


Note: Names of countries, territories or areas of geographical interest and their subdivisions are listed here
according to ISO country alpha-3-codes.






3.2 SEA AND MARITIME DISTANCES


Maritime distance is an estimated sea distance. It is obtained by summing sea distances on all


sea transport sections between two countries. When the connection is direct, maritime and sea


distances perfectly coincide.




Table 7 and Table 8 contain some basic statistics qualifying estimated maritime distances for


several countries or geographical groups of countries. Not surprisingly, countries in the Pacific region


are characterized by the largest mean and median values of maritime distance. Together with the fact


that countries in the region, including Australia and New Zealand, do not rank very well in terms of


average number of transhipments per connection, it makes the Pacific region the most remote one. On


the other extreme of the distribution are the United States, Canada and European countries. This


corroborates previous results on the average number of transhipments per connection. As a


consequence, the latter countries appear to be at the core of maritime connections. The Africa group


statistics are comparable to those of the European Union, although African countries do not present


any comparable performance in terms of number of transhipments per connection.




Changes over the 2006–2012 period have not been dramatic in most cases. The largest ones


are observed for countries in the Pacific region and for Asian countries.








Building a Dataset for Bilateral Maritime Connectivity 9


Table 7: Maritime distance (estimated): 2006


Mean Median SD CV Max Min


AUS 16 464 16 709 6 089 0.37 26 973 1 985


Africa 10 822 10 060 5 678 0.52 30 843 141


America 12 526 12 203 6 220 0.50 31 636 117


Asia 12 302 12 114 5 989 0.49 29 228 143


Canada 9 778 9 834 4 141 0.42 25 148 1 141


CHN 14 575 15 668 5 361 0.37 22 243 896


EUR 10 455 9 643 6 107 0.58 32 332 85


Europe 10 004 9 877 5 685 0.57 28 313 256


IND 10 899 11 119 5 712 0.52 24 746 941


JPN 15 017 15 972 5 801 0.39 24 007 1 241


NZL 16 899 17 074 6 010 0.36 28 423 2 280


Pacific 17 551 18 614 6 817 0.39 33 054 152


USA 9 685 9 688 4 692 0.48 26 197 165




Total 11 926 11 303 6 276 0.53 33 054 85




Note: Names of countries, territories or areas of geographical interest and their subdivisions are listed here
according to ISO country alpha-3-codes.



Table 8: Maritime distance (estimated): 2012


Mean Median SD CV Max Min


AUS 16 232 16 281 5 950 0.37 27 254 1 985


Africa 10 974 10 358 5 673 0.52 31 178 141


America 12 588 12 523 6 144 0.49 30 262 117


Asia 11 796 11 497 5 863 0.50 30 017 143


Canada 9 883 10 127 4 117 0.42 21 152 1 141


CHN 14 441 15 709 5 365 0.37 22 031 896


EUR 10 315 9 505 6 134 0.59 32 493 85


Europe 9 883 9 584 5 663 0.57 32 232 256


IND 10 965 11 025 5 873 0.54 24 461 941


JPN 15 288 15 907 6 158 0.40 25 374 1 241


NZL 17 531 17 438 6 611 0.38 29 515 2 280


Pacific 16 275 16 900 6 267 0.39 29 921 152


USA 9 451 9 173 4 487 0.47 21 630 165




Total 11 761 11 219 6 132 0.52 32 493 85


Note: Names of countries, territories or areas of geographical interest and their subdivisions are listed here
according to ISO country alpha-3-codes.




Average maritime distance for the Pacific region has fallen by more than 7 per cent and


median maritime distance by about 9 per cent. Average and median maritime distance for the Asian


countries group fell by about 5 per cent.




Overall, this trend can be considered positive. Although the number of direct connections has


decreased for many countries, a geographically wider distribution of major transhipment ports has


improved the options to connect trading partners with transhipments implying a lower distance to be


travelled by the traded container – albeit also requiring a larger number of transhipments.






10 POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES


Despite its exceptional impact on overall aggregate demand and trade, the financial crisis of


2008 does not seem to have deeply affected maritime distances. This may come as a surprise


considering the figures on the average number of transhipments reviewed previously. A clear exception


is New Zealand, whose mean and median maritime distance increased by more than 8 per cent


between 2008 and 2010 and have only marginally decreased since then.






Table 9: Variations in estimated maritime distances and number of transhipments


Variation Maritime distance (per cent) Number of transhipments (%)


2006–2012
>0 15 12.2


<0 16 10.2


2006–2008
>0 13.5 8


<0 14.4 12.3


2008–2010
>0 14 11.3


<0 17 9.5


2010–2012
>0 13 13


<0 12.6 7




A study of the variations in maritime distances and transhipments as reported in Table 9,


however, reveals features consistent to a large extent with the series’ average behaviour. Over the


whole period under investigation, 30 per cent of connections have varied in terms of maritime distance.


Among these 30 per cent, half of them lengthened and half of them shortened. Surprisingly enough, the


biennium following the financial crisis has been marked by a large share of shortened connections.


With regard to the number of transhipments, about 22 per cent of connections have varied over the


2006–2012 period. The number of transhipments necessary to connect two countries has increased for


12 per cent of connections and has decreased for about 10 per cent of them. The post-financial crisis


period has been characterized by an increasing share of connections necessitating a larger number of


transhipments.




The direct sea distance and the shortest connection distance with transhipment are by nature


strongly correlated. The maritime distance with transhipments, however, tends to increase with respect


to sea distance as the latter increases. The farther away two countries are from each other, the more


likely it is that they need more transhipments to trade with each other, and each transhipment implies


some deviation from the shortest (direct) route. Figure 2 reproduces this relationship for a selection of


years (left quadrant) and regions (right quadrant), which include the whole set of composing countries.


The relationship appears to be relatively stable during the period under observation. The pre-crisis


period is characterized to some extent by larger maximum maritime distances than the post-crisis


period.




There are some salient facts about regional relationships. Sample means are indicated by


vertical and horizontal dashed lines and the red curve connect fitted values based on a quadratic


approximation. Pacific countries were found to be characterized by relatively large maritime distances.


As shown in figure 2, this is a consequence of essentially larger sea distances from most trade


partners. As far as American countries are concerned, the quadratic fit is almost a linear fit. This is to a


large extent the reflection of a large number of direct connections to the United States, the


geographical configuration of the continent and the existence of the Panama Canal.




The whole set of relationships between direct sea distance and maritime distances with


transhipments presented above remain similar whether or not we include those country pairs with a


direct maritime connection. In the latter case, as mentioned previously, the two distances by definition


coincide.








Building a Dataset for Bilateral Maritime Connectivity 11


Figure 2: Direct sea distance and maritime distance (estimated) with transhipments




Note: The red line represents the linear fit of the relationship, the green line, its quadratic fit.






Figure 3: Direct sea distance and maritime distance (estimated)


50
00


10
00


0
15


00
0


50
00


10
00


0
15


00
0


10000 15000 20000 25000 10000 15000 20000 25000


2006 2008


2010 2012


Obs. Fitted values


D
ire


ct


Se
a



D


ist
a


n
ce


Maritime Distance (Estimated)


Graphs by Year


6
8


10
12


14
6


8
10


12
14


5 10 15 20


5 10 15 20 5 10 15 20


Africa America Asia


Europe Pacific


Obs. Fitted values


D
ire


ct


Se
a



D


ist
a


n
ce


Maritime Distance (Estimated)


Graphs by Continent


Note: The red line represents the quadratic fit of the relationship.


The question of whether maritime distances with transhipment and the associated number of


transhipments are correlated does not have an obvious answer. The linear and quadratic fit lines


reported in Figure 4 both suggest that the two measures are only weakly correlated. The right quadrant


reports similar fits when all direct connections are excluded. Even with that subsample, the two


distance measures remain only weakly correlated.




This result suggests that distance as such may not fully reflect the incidence of transport


costs, and it may have to be considered together with the number of transhipments in assessing the


impact of transport costs on bilateral exchanges.






50
00


10
00


0
15


00
0


20
00


0
Di


re
ct



Se


a


Di
st


a
n


ce


5000 10000 15000 20000
Maritime Distance (Estimated)





12 POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES


Figure 4: Maritime distance (estimated) and number of transhipments (country averages)


10
15


20
25


10
15


20
25


1 2 3 4 1 2 3 4


2006 2008


2010 2012


Obs. Fitted values
Fitted values


M
a


rit
im


e


D
ist


a
n


ce


(E
st


im
a


te
d)


Number of Transhipments


Graphs by Year


10
15


20
25


10
15


20
25


1 2 3 4 1 2 3 4


2006 2008


2010 2012


Obs. Fitted values
Fitted values


M
a


rit
im


e


D
ist


a
n


ce


(E
st


im
a


te
d)


Number of Transhipments


Graphs by Year


Note: The red line represents the linear fit of the relationship, the green line, its quadratic fit.




3.3 TRADE, MARITIME DISTANCE AND TRANSHIPMENTS


In the absence of extensive estimates of transport costs, distance has been used to proxy the


latter. However, previous results revealed that additional information on maritime transport costs may


be contained in the counting of transhipments necessary to move containers between any pair of


countries.




The intensive margin of trade


Figure 5 shows a scatter between total containerizable exports (period average) and the


estimated average maritime distance. The left quadrant refers to the whole sample while the right


quadrant refers to a sample without China, the United States, Japan and Germany. In the former case


the unconditional relationship between exports and maritime distance appears to be positive although


close to zero. When excluding the largest exporting countries, the unconditional relationship turns to


be negative, as expected.




Figure 6 illustrates the relationship between total containerizable exports and the number of


transhipments. Whether we include (left quadrant) or not (right quadrant) the largest exporters, the


unconditional relationship is clearly negative. In other words, bilateral trade tends to decrease with the


number of transhipments. Or, put differently, the direct connections tend to increase if demand (trade


in containerizable goods) so requires.







Building a Dataset for Bilateral Maritime Connectivity 13


Figure 5: Containerizable exports and maritime distance (estimated) for liner shipping


connections




ABWAFGAGOALB ARE ARGARM AT AUS
AUT


AZE BDI


BEL


BENBFA BGDGR BHR BHSBLRBLZ BOL
BRA


BRNWACAF


CANCHE


CHL


CHN


CIV CMRCODCOG OLMCPV CRICUBCYP
CZE


DEU


DJI DMA
DNK


D MDZA ECUEG


ESP


ESTETH
FIN


FJI


FRA
GBR


GEOGINGMBB QGRC RD GTMUY
HKG


HNDH V TI
HUN IDN


INDIRL


IRIRQISLISR


ITA


JAMJO


JPN


KEN KGZKHM KIRKNA


KOR


KWT LAOLBN LLB L LKALSOLTUUXLVAM RMD M DV


MEX


MHLKD IMLT MMR MNGOZRT USMWI


MYS


N NCLNER NIC


NLD


NO NRUNZLOMN PAK PAN PER
PHL


PLWPNG
POL


PRT PRY PYFQRUS RSAU SE


SGP


SLSLE SLVSO S PS SVKS N
SWE


S ZYSY TCDG


THA


TK TONTTTUN
TUR


TWN


TZAUGK URY


USA


V VEN
VNM


VUTWSME
ZAF


ZM Z0
50


0
10


00
15


00
To


ta
l E


xp
o


rts


(C
o


n
ta


in
e


riz
a


bl
e


)


5000 10000 15000 20000
Maritime Distance (Estimated)


ABWAFGAGOALB


ARE


ARG
ARM AT


AUS


AUT


AZE BDI


BEL


BENBFA
BGD


BGR BHR BHSBLRBLZ BOL


BRA


R BRNWACAF


CAN
CHE


CHL
CIV CMRCODCOG


COL
MCPV


CRI
CUBCYP


CZE


DJI DMA


DNK


DOMDZA ECU
EGY


ESP


ESTETH


FIN


FJI


FRA


GBR


GEOGINGMBB Q
GRC


RD GTMUY


HKG


HNDHRV HTI


HUN IDN


IND


IRL


IRNI QISL


ISR


ITA


JAMJOR KEN KGZKHM KIRKNA


KOR


KWT LAOLBN LBRLB L LKALSOLTU
LUX


LVAMARMDA M DV


MEX


MHLMKD LIMLT MMR MNGOZRT USMWI


MYS


N M NCLNER NIC


NLD


NOR


NRU
NZL


OMN
PAK


PAN PER


PHL


PLWPNG


POL


PRT


PRY PYFQ T


RUS


R


SAU
SE


SGP


SLSLE SLVSO S PS


SVK
SVN


SWE


S ZYSYR TCDTG


THA


TK TONTTO
TUN


TUR


TWN


TZAUG
KR


URYV VEN


VNM


VUTWSMYE


ZAF


ZM Z0
10


0
20


0
30


0
To


ta
l E


xp
o


rts


(C
o


n
ta


in
e


riz
a


bl
e


)


5000 10000 15000 20000
Maritime Distance (Estimated)




ABW


AFG
AGOALB


ARE
ARG


ARM ATG


AUS
AUT


AZE


BDI


BEL


BENBFA


BGD
BGR


BHR


BHS


BLR


BLZ
BOL


BRA


BRB BRN


BWA


CAF


CANCHE


CHL


CHN


CIV
CMRCOD


COG


COL


COMCPV


CRI


CUB
CYP


CZE


DEU


DJI
DMA


DNK


DOM


DZA


ECUEGY


ESP


EST


ETH


FIN


FJI


FRA


GAB


GBR


GEO
GH


GIN
GMB


GNB


GNQ


GRC


GRD


GTM


GUY


HKG


HNDHRV


HTI


HUN IDN
INDIRL


IRN


IRQ


ISL


ISR


ITA


JAM


JOR


JPN


KEN


KGZ


KHM


KIR


KNA


KOR


KWT


LAO


LBN


LBR


LBY


LCA


LKA


LSO


LTULUX
LVA


MAR


MDA MDG


MDV


MEX


MHL


MKD


MLI


MLT
MMR


MNG


MOZ


MRT


MUS
MWI


MYS


NAM


NCL


NER


NGA NIC


NLD


NOR


NRU


NZL


OMN


PAK


PAN
PER


PHL


PLW


PNG


POL
PRT


PRY
PYF


QAT


RU


RWA


SAU


SEN


GP


SLB


SLE


SLV


SOM


STP


S R


SVK
SVN


SWE


SWZ
SYC


SYR


TCD
TGO


THA


TKM


TON


TTO


TUN


TUR


TWN


TZA
UGA


KR


URY


USA


VCT


VEN


VNM


VUT


WSM


YEM


ZAF


ZMB ZWE


-
10


-
5


0
5


10
To


ta
l E


xp
o


rts


(C
o


n
ta


in
e


riz
a


bl
e


)


9 9.2 9.4 9.6 9.8
Maritime Distance (Estimated)


ABW


AFG
AGOALB


ARE
ARG


ARM
ATG


AUS
AUT


AZE


BDI


BEL


BENBFA


BGD
BGR


BHR


BHS


BLR


BLZ


BOL


BRA


BRB BRN


BWA


CAF


CANCHE


CHL


CIV
CMRCOD


COG


COL


COMCPV


CRI


CUB
CYP


CZE


DJI
DMA


DNK


DOM


DZA


ECU
EGY


ESP


EST


ETH


FIN


FJI


FRA


GAB


GBR


GEO
GH


GIN
GMB


GNB


GNQ


GRC


GRD


GTM


GUY


HKG


HNDHRV


HTI


HUN IDN
INDIRL


IRN


IRQ


ISL


ISR


ITA


JAM


JOR KEN


KGZ


KHM


KIR


KNA


KOR


KWT


LAO


LBN


LBR


LBY


LCA


LKA


LSO


LTULUX


LVA


MAR


MDA MDG


MDV


MEX


MHL


MKD


MLI


MLT


MMR


MNG


MOZ


MRT


MUS
MWI


MYS


NAM


NCL


NER


NGA
NIC


NLD


NOR


NRU


NZL


OMN


PAK


PAN
PER


PHL


PLW


PNG


POL
PRT


PRY
PYF


QAT


RU


RWA


SAU


SEN


GP


SLB


SLE


SLV


SOM


STP


S R


SVK
SVN


SWE


SWZ
SYC


SYR


TCD


TGO


THA


TKM


TON


TTO


TUN


TUR


TWN


TZA
UGA


KR


URY


VCT


VEN


VNM


VUT


WSM


YEM


ZAF


ZMB
ZWE


-
10


-
5


0
5


To
ta


l E
xp


o
rts



(C


o
n


ta
in


e
riz


a
bl


e
)


9 9.2 9.4 9.6 9.8
Maritime Distance (Estimated)




Note: Values in upper quadrants are in levels, and values in lower quadrants are in natural logs. The red line
represents the linear fit of the relationship, the green line, its quadratic fit.






Figure 6: Containerizable exports and number of transhipments




ABWAFGAGO ALBARE ARG ARMATGAUS
AUT


AZEBDI


BEL


BENBFA BGDBGRBHRBHS BLR BLZOL
BRA


BRB BRNBW CAF


CANCHE


CHL


CHN


CIV CMR CODCOGCOL COMCPVCRI UBCYP
CZE


DEU


DJI DMA
DNK


DOM DE UEGY


ESP


ESTTH
FIN


FJI


FRA


A


GBR


GEOGHA IN GMB GNBGNQGR G DGTM GUY
HKG


DHRV HTI
HUNI


IND IRL


IR IRQISLISR


ITA


JA JO


JPN


KENKGZ KHMKIRKNA


KOR


KWTLL N LBRLB LCALKA L O LTULUX LVAMA MDMDG MDV


MEX


MHLMKD MLIM T RMN M Z TMUSMWI


MYS


NAL ERN A IC


NLD


NOR NRUNZO NPAKP NPE
PHL


P WPN
POL


PRT PRYPYF QAUS R ASAU SE


SGP


SLBSLE SLV S MSTP SSVKSVN
SWE


S SYCS RTCT


THA


T M TONT UN
T R


TWN


TZ UGA KRURY


USA


VCE
VNM


VUT WSYEMZAF ZMBE0
50


0
10


0
15


00
To


ta
l E


xp
o


rts


(C
o


n
ta


in
e


riz
a


bl
e


)


0 1 2 3
Number of Transhipments


ABWAFGAGO ALB


ARE
ARG


ARMATG


AUS


AUT


AZEBDI


BEL


BENBFA
BGDBGRBHRBHS BLR BLZOL


BRA


BRB BRNBW CAF


CAN
CHE


CHLCIV CMR CODCOGCOL COMCPV
CRI


UBCYP


CZE


DJI DMA


DNK


DOM DECUEGY


ESP


ESTTH


FIN


FJI


FRA


A


GBR


GEOGHA IN GMB GNBGNQGR G DGTM GUY


HKG


HNDHRV HTI


HUNIDN


IND
IRL


IRN IRQISL


ISR


ITA


JA JOKENKGZ KHMKIRKNA


KOR


KWTLL N LBRLB LCALKA L O LTULUX LVAMAR MDMDG MDV


MEX


MHLMKD MLIMLT RMNG M Z TMUSMWI


MYS


NAL ERNGA IC


NLD


NOR
NRU


NZ
O NPAKP NPER


PHL


P WPN


POL


PRT


PRYPYF QAT


US


R A
SAU


SE


SGP


SLBSLE SLV S MSTP S
SVK


SVN


SWE


S SYCSYRTCT


THA


T M TONT TUN


TUR


TWN


TZ UGA KRU Y VCEN


VNM


VUT WSMYEM


ZAF


ZMBE0
20


0
40


0
To


ta
l E


xp
o


rts


(C
o


n
ta


in
e


riz
a


bl
e


)


0 1 2 3
Number of Transhipments




Note: The red line represents the linear fit of the relationship, the green line, its quadratic fit.






On the contrary, direct connections are likely to be positively associated with exports. Figure 7


reports the relationship between direct connections and containerizable exports. Unsurprisingly, the


association is clearly positive. Once again, the relationship does not seem to be driven by outliers. It


remains clearly positive even after outliers such as the largest exporters are excluded from the sample


(right quadrant).










14 POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES


Figure 7: Containerizable exports and direct connections




0
20


0
40


0
To


ta
l E


xp
o


rts


(C
o


n
ta


in
e


riz
a


bl
e


)


0 20 40 60 80 100
Direct Connections




Note: The red line represents the linear fit of the relationship, the green line, its quadratic fit.






The extensive margin of trade


Previous graphs were focused on active trade relationships. However, about one third of


containerizable trade flows among countries in our sample are zero. Transports costs and their


connectivity component may be good predictors of trade patterns at its extensive margin. This is


visible in Figure 8. The number of direct connections affects the incidence of zero trade (left quadrant).


Countries characterized by a larger number of direct connections show a smaller number of zero trade


flows. The right quadrant of Figure 8 reveals that as the average number of transhipments necessary to


connect to any country increases, the incidence of zero trade flows also increases. Without talking


about causality, the creation of direct connections could help remote economies promote their exports.






Figure 8: Zero trade and maritime connectivity




0
50


10
0


15
0


0
50


10
0


15
0


0 50 100 0 50 100


2006 2008


2010 2012


Obs. Fitted values
Fitted values


Ze
ro



Tr


a
de


Direct Connections


Graphs by Year


0
50


10
0


15
0


0
50


10
0


15
0


1 2 3 4 1 2 3 4


2006 2008


2010 2012


Obs. Fitted values
Fitted values


Ze
ro



Tr


a
de


Number of Transhipments


Graphs by Year


Note: The red line represents the linear fit of the relationship, the green line, its quadratic fit.






Trade imbalances


About 20 per cent of trade relationships are unilateral. This means that for about 20 per cent of


the country pairs represented in the data, a zero containerizable trade flow in one direction is


associated with a positive trade flow in the opposite direction. This is an extreme illustration of


asymmetric trade flows. However, all bilateral trade flows are asymmetric to some extent.






0
50


0
15


00
To


ta
l E


xp
o


rts


(C
o


n
ta


in
e


riz
a


bl
e


)


0 20 40 60 80 100
Direct Connections







Building a Dataset for Bilateral Maritime Connectivity 15


Figure 9 reports for a selection of years the relationship between a measure of country pair


trade unbalance, the number of transhipments to connect the country pair and the corresponding


maritime distance, respectively. Trade imbalances are measures by the absolute value of the difference


(absolute) between the two trade flows. Nothing really significant comes out of a basic graphical


analysis. If at all related, the relationship could be only slightly negative. Trade imbalances would tend


to diminish as the number of transhipments and the maritime distance increase.






Figure 9: Trade imbalances and connectivity




0
20


40
60


80
0


20
40


60
80


1 3 5 7 1 3 5 7


2006 2008


2010 2012


Tr
a


de


Im
ba


la
n


ce
s


(A
bs


o
lu


te


Va
lu


e
)


Number of Transhipments
Graphs by Year


0
20


40
60


80
0


20
40


60
80


0 10000 20000 30000 0 10000 20000 30000


2006 2008


2010 2012


Tr
a


de


Im
ba


la
n


ce
s


(A
bs


o
lu


te


Va
lu


e
)


Maritime Distance (Estimated)
Graphs by Year








4. APPLICATIONS AND FUTURE RESEARCH


Despite the importance of trade costs as drivers of the geographical pattern of economic


activity, global value chains, and of exchanges of merchandise goods between countries, most


contributions to their understanding remain piecemeal.




Traditionally sea distance is assumed to be among the main determinants of freight rates and


thus also of the trade competitiveness of countries. Findings by Wilmsmeier and Hoffmann (2008)


based on a sample of 189 freight rates of one company for the Caribbean confirm to some extent the


general positive correlation between distance and freight rates. However, sea distance explains only


one fifth of the variance of the freight rate. Other possible determinants of trade competitiveness are


transport connectivity, defined as the access to regular and frequent transport services and the level of


competition in the service supply. The basic set of variables to account for transport costs are sea


(maritime) distance, various aspects of liner shipping connectivity, trade balance of containerizable


goods, various aspects of port infrastructure endowment and the countries’ general level of


development. As mentioned previously, Wilmsmeier and Hoffmann (2008) also show that trade routes


with only indirect services (i.e. including transhipments) induce higher transport costs. Unconditional


correlations between our two measures and trade of containerizable goods presented in the previous


section appear to be supportive of such conclusions.




In this context, the definition of the number of transhipments necessary to connect any


country pair and the computation of the corresponding effective maritime distance for a sample of 178


countries during a six-year period is a clear contribution to the empirics of trade. Our two variables


could be of immediate use in the analysis of transport costs and their implications for bilateral trade.


However, a clear causal relationship may be difficult to identify, as there are most probably serious


endogeneity issues related to either reverse causality or variables or both. Further research is


necessary and will be forthcoming in a companion paper.




Connectivity has become an increasingly popular research topic. However, a clearly


established bilateral connectivity index for shipping is still lacking. Our two variables can contribute to


the establishment of such an index. The latter could be based on the combination of our two





16 POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES


constructed variables and of some liner shipping connectivity aspects. This procedure is in line with a


recent tentative index building on UNCTAD’s country-level Liner Shipping Connectivity Index (LSCI)


and would be called LSBCI (Liner Shipping Bilateral Connectivity Index). Generally speaking, four sets


of components should be considered for the development of a bilateral index. First, the number of


companies providing direct services between two countries should be represented. A simple version of


this component would be a dummy variable which assumes the value 1 if a direct service exists at all,


and 0 if not. A more sophisticated version would include the number of transhipments necessary to


connect any pair of countries, as computed in this paper. Second, the number of common


connections between any country A and any country B should also be included. A simple version of


this component would be a dummy variable, which assumes the value 1 if exists an option to connect


the two countries with one transhipment, and 0 if not. By the same token, the number of second-level


connections could be generated, i.e. how many options there are to get from country A to country B


with two transhipments. Third, combinations of both countries’ LSCI, such as the product, or the


geometric average of both countries’ index should be considered. The Index already includes five


components, notably the number of ships, their TEU capacity, the size of the largest ship, the number


of companies and the number of services. Finally, data on vessel deployment with transhipment


options included should be used. Even for pairs of countries without a direct connection, it is possible


to generate what are the best connections between them under specific criteria, such as the number of


companies in the market or the largest ships deployed on the different legs of a connection with one or


more transhipments. This represents an immediate application of the algorithm developed previously


with an additional cost reference. Instead of solely considering the sea distance, we would also


consider the number of companies or the largest vessel size deployed in identifying the shortest path.


The development of the country-level Liner Shipping Connectivity Index (LSCI) has shown to be useful


for policymakers and researchers. It can help to illustrate trends in a country’s connectivity to the


global liner shipping network. The development a similar type of index for pairs of countries would


certainly enlarge the scope of the country-level LSCI.








Building a Dataset for Bilateral Maritime Connectivity 17


ANNEX: LIST OF COUNTRY ALPHA-3-CODES


Country


code


Country


name


Country


code


Country


name



ABW Aruba GHA Ghana
AFG Afghanistan GIN Guinea
AGO Angola GMB Gambia
ALB Albania GNB Guinea-Bissau
ARE United Arab Emirates GNQ Equatorial Guinea
ARG Argentina GRC Greece
ARM Armenia GRD Grenada
ASM American Samoa GTM Guatemala
ATG Antigua and Barbuda GUM Guam
AUS Australia GUY Guyana
AUT Austria HKG China, Hong Kong Special Administrative Region
AZE Azerbaijan HND Honduras
BDI Burundi HRV Croatia
BEL Belgium HTI Haiti
BEN Benin HUN Hungary
BFA Burkina Faso IDN Indonesia
BGD Bangladesh IND India
BGR Bulgaria IRL Ireland
BHR Bahrain IRN Iran (Islamic Republic of)
BHS Bahamas IRQ Iraq
BLR Belarus ISL Iceland
BLZ Belize ISR Israel
BMU Bermuda ITA Italy
BOL Bolivia (Plurinational State of) JAM Jamaica
BRA Brazil JOR Jordan
BRB Barbados JPN Japan
BRN Brunei Darussalam KAZ Kazakhstan
BTN Bhutan KEN Kenya
BWA Botswana KGZ Kyrgyzstan
CAF Central African Republic KHM Cambodia
CAN Canada KIR Kiribati
CHE Switzerland KNA Saint Kitts and Nevis
CHL Chile KOR Republic of Korea
CHN China KWT Kuwait
CIV Côte d'Ivoire LAO Lao People's Democratic Republic
CMR Cameroon LBN Lebanon
COD Democratic Republic of the Congo LBR Liberia
COG Congo LBY Libya
COK Cook Islands LCA Saint Lucia
COL Colombia LKA Sri Lanka
COM Comoros LSO Lesotho
CPV Cabo Verde LTU Lithuania
CRI Costa Rica LUX Luxembourg
CUB Cuba LVA Latvia
CYM Cayman Islands MAR Morocco
CYP Cyprus MDA Republic of Moldova
CZE Czech Republic MDG Madagascar
DEU Germany MDV Maldives
DJI Djibouti MEX Mexico
DMA Dominica MHL Marshall Islands
DNK Denmark MKD The former Yugoslav Republic of Macedonia
DOM Dominican Republic MLI Mali
DZA Algeria MLT Malta
ECU Ecuador MMR Myanmar
EGY Egypt MNG Mongolia
ERI Eritrea MOZ Mozambique
ESP Spain MRT Mauritania
EST Estonia MSR Montserrat
ETH Ethiopia MUS Mauritius
FIN Finland MWI Malawi
FJI Fiji MYS Malaysia
FRA France NAM Namibia
GAB Gabon NCL New Caledonia


GBR
United Kingdom of Great Britain and Northern
Ireland NER Niger


GEO Georgia NGA Nigeria


…/…










18 POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES





Country


code


Country


name



NIC Nicaragua
NLD Netherlands
NOR Norway
NPL Nepal
NRU Nauru
NZL New Zealand
OMN Oman
PAK Pakistan
PAN Panama
PER Peru
PHL Philippines
PLW Palau
PNG Papua New Guinea
POL Poland
PRT Portugal
PRY Paraguay
PYF French Polynesia
QAT Qatar
ROU Romania
RUS Russian Federation
RWA Rwanda
SAU Saudi Arabia
SDN Sudan
SEN Senegal
SGP Singapore
SLB Solomon Islands
SLE Sierra Leone
SLV El Salvador
SOM Somalia
STP Sao Tome and Principe
SUR Suriname
SVK Slovakia
SVN Slovenia
SWE Sweden
SWZ Swaziland
SYC Seychelles
SYR Syrian Arab Republic
TCD Chad
TGO Togo
THA Thailand
TJK Tajikistan
TKM Turkmenistan
TON Tonga
TTO Trinidad and Tobago
TUN Tunisia
TUR Turkey
TZA United Republic of Tanzania
UGA Uganda
UKR Ukraine
URY Uruguay
USA United States of America
UZB Uzbekistan
VCT Saint Vincent and the Grenadines
VEN Venezuela (Bolivarian Republic of)
VNM Viet Nam
VUT Vanuatu
WSM Samoa
YEM Yemen
ZAF South Africa
ZMB Zambia
ZWE Zimbabwe










Building a Dataset for Bilateral Maritime Connectivity 19




REFERENCES


Arvis JF and Shepherd B (2011). The air connectivity index: Measuring integration in the global air
transport network. Policy Research Working Paper 5722. World Bank.


Arvis JF, Shepherd B, Duval Y, and Utoktham C (2013). Trade costs and development: a new data set.


Economic Premise (104). http://siteresources.worldbank.org/EXTPREMNET/Resources/
EP104.pdf.


Bernhofen DM, El-Sahli Z and Kneller R (2013). Estimating the effects of the container revolution on


world trade. CESifo Working Paper No. 4136. CESifo, Centre for Economic Studies and Ifo
Institute.




Caschili S and Romana Medda F (2012). A review of the maritime container shipping industry as a
complex adaptive system. Interdisciplinary Description of Complex Systems. 10(1):1–15.


CEPII (2013). Centre de recherche français dans le domaine de l’économie internationale. GeoDist.
http://www.cepii.fr/anglaisgraph/bdd/distances.htm.


Cormen TH, Leiserson CE, Rivest RL, Stein, Clifford (2001). Section 24.3: Dijkstra's algorithm. In:


Introduction to Algorithms (Second edition.). MIT Press and McGraw-Hill. pp. 595–601. ISBN
0-262-03293-7.



Duval Y (2011). UN ESCAP: Trade Facilitation in Asia and the Pacific: Which Policies and Measures


affect Trade Costs the Most? Bangkok. http://www.unescap.org/tid/publication/swp111.pdf.


Ducruet C and Notteboom M (2010). The Worldwide Maritime Network of Container Shipping: Spatial


Structure and Regional Dynamics.
GaWC Research Bulletin 364.


Dijkstra, E W (1959). A note on two problems in connexion with graphs. Numerische Mathematik.


1:269–271.


The Economist (2013). The humble hero. 18 May.


Hoffmann J (2012). Corridors of the sea: An investigation into liner shipping connectivity. Les Corridors
de Transport. Les océanides.




Hu Y and Zhu D (2008). Empirical analysis of the worldwide maritime transportation network. Physica
A. 388 (1061–2071). June.


Hummels D and Schaur G (2012). Time as a Trade Barrier. NBER Working Paper 17758, January.
http://www.nber.org/papers/w17758.pdf.


International Civil Aviation Organization (2013). Enhancement of Air Transport Connectivity and
Facilitation. Worldwide Air Transport Conference. Montreal.


Kaluza P, Kolazsch A, Gastner MT and Blasius B (2010). The complex network of global cargo ship


movements. Journal of the Royal Society Interface. 7(48):1093–1103.


Mandryk W (2009). Measuring Global Seaborne Trade. Presented at the annual meeting of the


International Maritime Statistics Forum. New Orleans, 4–6 May 2009.
www.imsf.info/papers/NewOrleans2009/documents/Wally_Mandryk_LMIU_IMSF09.pdf.


UNCTAD (2008) Transport Newsletter. No. 38, March. http://www.unctad.org/en/docs/


sdtetlbmisc20081_en.pdf.


________ (2011). Review of Maritime Transport 2011. United Nations. Geneva.


http://www.unctad.org/rmt.





20 POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES


________ (2012). Review of Maritime Transport 2012. United Nations. Geneva.
http://www.unctad.org/rmt.


________ (2013). Review of Maritime Transport 2012. United Nations. Geneva.
http://www.unctad.org/rmt.


________ (2013). UNCTAD STAT online. http://unctadstat.unctad.org/TableViewer/
tableView.aspx?ReportId=92.


UNECLAC (1998). Concentration in Liner Shipping – its causes and impacts for ports and shipping


services in developing countries, Santiago. http://www.eclac.org/publicaciones/


xml/5/5175/LC_G.2027.pdf.


UNESCAP (2008). Gravity models: Theoretical foundations and related estimation issues. Presentation


made at ARTNetCapacity-Building Workshop for Trade Research. Phnom Penh, Cambodia.
2–6 June 2008. http://www.unescap.org/tid/artnet/mtg/cbcam_d2s3.pdf


Wilmsmeier G and Hoffmann J (2008). Liner shipping connectivity and port infrastructure as


determinants of freight rates in the Caribbean. Maritime Economics and Logistics. 130–151.


Wilmsmeier G, Hoffmann J and Sanchez R (2006). The impact of port characteristics on international


maritime transport costs. In: Cullinane K and Talley W, eds. Port Economics. Vol. 16. Elsevier.


Wilmsmeier G and Notteboom T (2010). Determinants of liner shipping network configuration: A two-


region comparison. Geojournal. DOI:10.1007/s10708-009-9333-2.


Yap WY and Notteboom T (2011). Dynamics of liner shipping service scheduling and their impact on


container port competition. Maritime Policy and Management. 38(5):471–485.








Building a Dataset for Bilateral Maritime Connectivity 21




UNCTAD Study Series


POLICY ISSUES IN INTERNATIONAL TRADE
AND COMMODITIES














No. 30 Sam Laird, David Vanzetti and Santiago Fernández de Córdoba, Smoke and mirrors:
Making sense of the WTO industrial tariff negotiations, 2006, Sales No.
E.05.II.D.16.




No. 31 David Vanzetti, Santiago Fernandez de Córdoba and Veronica Chau, Banana split:
How EU policies divide global producers, 2005, 27 p. Sales No. E.05.II.D.17.




No. 32 Ralf Peters, Roadblock to reform: The persistence of agricultural export subsidies,
2006, 43 p. Sales No. E.05.II.D.18.




No. 33 Marco Fugazza and David Vanzetti, A South–South survival strategy: The potential
for trade among developing countries, 2006, 25 p.






No. 34 Andrew Cornford, The global implementation of Basel II: Prospects and outstanding
problems, 2006, 30 p.




No. 35 Lakshmi Puri, IBSA: An emerging trinity in the new geography of international
trade, 2007, 50 p.




No. 36 Craig VanGrasstek, The challenges of trade policymaking: Analysis, communication
and representation, 2008, 45 p.




No. 37 Sudip Ranjan Basu, A new way to link development to institutions, policies and
geography, 2008, 50 p.




No. 38 Marco Fugazza and Jean-Christophe Maur, Non-tariff barriers in computable general
equilibrium modelling, 2008, 25 p.




No. 39 Alberto Portugal-Perez, The costs of rules of origin in apparel: African preferential
exports to the United States and the European Union, 2008, 35 p.




No. 40 Bailey Klinger, Is South–South trade a testing ground for structural
transformation?, 2009, 30 p.




No. 41 Sudip Ranjan Basu, Victor Ognivtsev and Miho Shirotori, Building trade-relating
institutions and WTO accession, 2009, 50 p.




No. 42 Sudip Ranjan Basu and Monica Das, Institution and development revisited: A
nonparametric approach, 2010, 26 p.






22 POLICY ISSUES IN INTERNATIONAL TRADE AND COMMODITIES


No. 43 Marco Fugazza and Norbert Fiess, Trade liberalization and informality: New stylized
facts, 2010, 45 p.




No. 44 Miho Shirotori, Bolormaa Tumurchudur and Olivier Cadot, Revealed factor intensity
indices at the product level, 2010, 55 p.




No. 45 Marco Fugazza and Patrick Conway, The impact of removal of ATC Quotas on
international trade in textiles and apparel, 2010, 50 p.




No. 46 Marco Fugazza and Ana Cristina Molina, On the determinants of exports survival,
2011, 40 p.




No. 47 Alessandro Nicita, Measuring the relative strength of preferential market access,
2011, 30 p.




No. 48 Sudip Ranjan Basu and Monica Das, Export structure and economic performance in
developing countries: Evidence from nonparametric methodology, 2011, 58 p.




No. 49 Alessandro Nicita and Bolormaa Tumurchudur-Klok, New and traditional trade flows
and the economic crisis, 2011, 22 p.




No. 50 Marco Fugazza and Alessandro Nicita, On the importance of market access for trade,
2011, 35 p.




No. 51 Marco Fugazza and Frédéric Robert-Nicoud, The ‘Emulator Effect’ of the Uruguay
round on United States regionalism, 2011, 45 p.




No. 52 Sudip Ranjan Basu, Hiroaki Kuwahara and Fabien Dumesnil, Evolution of non-tariff
measures: Emerging cases from selected developing countries, 2012, 38p.




No. 53 Alessandro Nicita and Julien Gourdon, A preliminary analysis on newly collected data
on non-tariff measures, 2013, 31 p.




No. 54 Alessandro Nicita, Miho Shirotori and Bolormaa Tumurchudur Klok, Survival analysis
of the exports of least developed countries: The role of comparative advantage,
2013, 25 p.




No. 55 Alessandro Nicita, Victor Ognivtsev and Miho Shirotori, Global supply chains: Trade
and Economic policies for developing countries, 2013, 33 p.




No. 56 Alessandro Nicita, Exchange rates, international trade and trade policies, 2013, 29 p.


No. 57 Marco Fugazza, The economics behind non-tariff measures: Theoretical insights and
empirical evidence, 2013, 33 p.




No. 58 Marco Fugazza and Alain McLaren, Market access, export performance and
survival: Evidence from Peruvian firms, 2013, 39 p.




No. 59 Patrick Conway, Marco Fugazza and M. Kerem Yuksel, Turkish enterprise-level
response to foreign trade liberalization: The removal of agreements on textiles and
clothing quotas, 2013, 54 p.




No. 60 Alessandro Nicita and Valentina Rollo, Tariff preferences as a determinant for
exports from Sub-Saharan Africa, 2013, 30 p.








Building a Dataset for Bilateral Maritime Connectivity 23


No. 61 Marco Fugazza, Jan Hoffmann and Rado Razafinombana, Building a dataset for
bilateral maritime connectivity, 2013, 31 p.


























































































Copies of the UNCTAD study series Policy Issues in International Trade and Commodities may be
obtained from the Publications Assistant, Trade Analysis Branch, Division on International Trade in
Goods and Services, and Commodities, United Nations Conference on Trade and Development,
Palais des Nations, CH-1211 Geneva 10, Switzerland (Tel: +41 22 917 4644).
These studies are available at http://unctad.org/tab.






Since 1999, the Trade Analysis Branch of the Division on International Trade in Goods and
Services, and Commodities of UNCTAD has been carrying out policy­oriented analytical work
aimed at improving the understanding of current and emerging issues in international trade and
development. In order to improve the quality of the work of the Branch, it would be useful to
receive the views of readers on this and other similar publications. It would therefore be greatly
appreciated if you could complete the following questionnaire and return to:


Trade Analysis Branch, DITC
Rm. E­8065


United Nations Conference on Trade and Development
Palais des Nations, CH­1211 Geneva 10, Swi+ erland


(Fax: +41 22 917 0044; E­mail: tab@unctad.org)


1. Name and address of respondent (optional):


2. Which of the following describes your area of work?


Government Public enterprise


Private enterprise institution Academic or research


International organization Media


Not­for­profi t organization Other (specify) _________________


3. In which country do you work? _________________________________________


4. Did you fi nd this publication Very useful Of some use Li8 le use
to your work?


5. What is your assessment of the contents of this publication?
Excellent Good Adequate Poor


6. Other comments:


QUESTIONNAIRE


UNCTAD Study series on


POLICY ISSUES IN INTERNATIONAL TRADE


AND COMMODITIES


(Study series no. 61: Building a dataset
for bilateral maritime connectivity)


Readership Survey




Login