Page 1
13 0 ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT
TRANSPORT AND INTERNATIONAL TRADE With the removal of many trade policy barriers, further international economic integration depends largely on the reduction of trade costs originating in the transport sector.
Background papers were provided by David Hummels (Purdue University), Anthony Venables (London School of Economics and Centre for Economic Policy Research) as well as Harry Broadman and John S. Wilson (World Bank).
TRANSPORT AND INTERNATIONAL TRADE
The Round Table discussion focused on the structure and development of international transport costs over the past decades and the benefits to be expected from investment in international transport facilities and the reduction of the costs of crossing borders.
130
R O U N D TA B L E
(75 2006 13 1 P1) ISBN 92-821-1338-8
-:HSTCSC=VVXX]\:
2006
C E N T R E
16:26
R E S E A R C H
27/07/06
T R A N S P O R T
fullTab130GB
EUROPEAN CONFERENCE OF MINISTERS OF TRANSPORT
TRANSPORT AND INTERNATIONAL TRADE
ROUND TABLE
13 0
TRANSPORT RESEARCH CENTRE
REPORT OF THE ONE HUNDRED AND THIRTIETH ROUND TABLE ON TRANSPORT ECONOMICS held in Paris on 21st-22nd October 2004 on the following topic:
TRANSPORT AND INTERNATIONAL TRADE
ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT
EUROPEAN CONFERENCE OF MINISTERS OF TRANSPORT
ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT The OECD is a unique forum where the governments of 30 democracies work together to address the economic, social and environmental challenges of globalisation. The OECD is also at the forefront of efforts to understand and to help governments respond to new developments and concerns, such as corporate governance, the information economy and the challenges of an ageing population. The Organisation provides a setting where governments can compare policy experiences, seek answers to common problems, identify good practice and work to co-ordinate domestic and international policies. The OECD member countries are: Australia, Austria, Belgium, Canada, the Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Korea, Luxembourg, Mexico, the Netherlands, New Zealand, Norway, Poland, Portugal, the Slovak Republic, Spain, Sweden, Switzerland, Turkey, the United Kingdom and the United States. The Commission of the European Communities takes part in the work of the OECD. OECD Publishing disseminates widely the results of the Organisation’s statistics gathering and research on economic, social and environmental issues, as well as the conventions, guidelines and standards agreed by its members.
This work is published on the responsibility of the Secretary-General of the OECD. The opinions expressed and arguments employed herein do not necessarily reflect the official views of the Organisation or of the governments of its member countries.
© OECD 2006 No reproduction, copy, transmission or translation of this publication may be made without written permission. Applications should be sent to OECD Publishing:
[email protected] or by fax (33 1) 45 24 13 91. Permission to photocopy a portion of this work should be addressed to the Centre français d'exploitation du droit de copie, 20, rue des Grands-Augustins, 75006 Paris, France (
[email protected]).
EUROPEAN CONFERENCE OF MINISTERS OF TRANSPORT (ECMT)
The European Conference of Ministers of Transport (ECMT) is an inter-governmental organisation established by a Protocol signed in Brussels on 17 October 1953. It comprises the Ministers of Transport of 43 full Member countries: Albania, Armenia, Austria, Azerbaijan, Belarus, Belgium, Bosnia-Herzegovina, Bulgaria, Croatia, the Czech Republic, Denmark, Estonia, Finland, France, FRY Macedonia, Georgia, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Liechtenstein, Lithuania, Luxembourg, Malta, Moldova, Netherlands, Norway, Poland, Portugal, Romania, Russia, Serbia and Montenegro, Slovakia, Slovenia, Spain, Sweden, Switzerland, Turkey, Ukraine and the United Kingdom. There are seven Associate member countries (Australia, Canada, Japan, Korea, Mexico, New Zealand and the United States) and one Observer country (Morocco). The ECMT is a forum in which Ministers responsible for transport, and more specifically the inland transport sector, can co-operate on policy. Within this forum, Ministers can openly discuss current problems and agree upon joint approaches aimed at improving the utilization and at ensuring the rational development of European transport systems of international importance. At present, ECMT has a dual role. On one hand it helps to create an integrated transport system throughout the enlarged Europe that is economically efficient and meets environmental and safety standards. In order to achieve this, it is important for ECMT to help build a bridge between the European Union and the rest of the European continent at a political level. On the other hand, ECMT's mission is also to develop reflections on long-term trends in the transport sector and to study the implications for the sector of increased globalisation. The activities in this regard have recently been reinforced by the setting up of a New Joint OECD/ECMT Transport Research Centre.
Also available in french under the title: Transport et commerce international
Further information about the ECMT is available on Internet at the following address: www.cemt.org © ECMT 2006 – ECMT Publications are distributed by: OECD Publications Service, 2, rue André Pascal, 75775 PARIS CEDEX 16, France
TABLE OF CONTENTS -
5
TABLE OF CONTENTS
INTRODUCTORY REPORTS Transportation Costs and Trade over Time – by D. HUMMELS (United States) .................. 7 1. 2. 3. 4. 5.
Introduction............................................................................................................................ 11 The import wedge: prices at home and abroad...................................................................... 12 The sourcing wedge ............................................................................................................... 19 Cross-commodity variation: transport intensity.................................................................... 21 Conclusions ............................................................................................................................ 24
Trade and Transport Facilitation: European Accession and Capacity Building Priorities – by J. WILSON, X. LUO and H. BROADMAN (World Bank, United States) ...................... 27 Abstract.......................................................................................................................................... 31 1. Introduction............................................................................................................................ 32 2. Trade and regional integration in Europe: empirical analysis for capacity building............. 33 3. Benchmarking the relationship between economic development and trade facilitation ........ 37 4. Externalities from investment in trade facilitation................................................................. 42 5. Concluding remarks: informing capacity-building policy .................................................... 49
Infrastructure, Trade Costs and the Gains from International Trade – by A. VENABLES (United Kingdom) ....................................................................................... 57 1. 2. 3. 4. 5.
Introduction............................................................................................................................ 61 Trade costs and real income: analysis ................................................................................... 61 Trade and real income: empirics ........................................................................................... 71 Trade costs, trade flows and infrastructure ............................................................................ 78 Concluding comments............................................................................................................ 81
SUMMARY OF DISCUSSIONS (Round Table debate on reports) ................................................................................................... 87
LIST OF PARTICIPANTS............................................................................................................... 103
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRANSPORTATION COSTS AND TRADE OVER TIME
TRANSPORTATION COSTS AND TRADE OVER TIME
David HUMMELS Purdue University West Lafayette, Indiana United States
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
-7
TRANSPORTATION COSTS AND TRADE OVER TIME
-9
TRANSPORTATION COSTS AND TRADE OVER TIME
SUMMARY
1.
INTRODUCTION ........................................................................................................................ 11
2.
THE IMPORT WEDGE: PRICES AT HOME AND ABROAD................................................ 12 2.1. Modal prices.......................................................................................................................... 15
3.
THE SOURCING WEDGE.......................................................................................................... 19 3.1. The role of distance............................................................................................................... 19 3.2. Scale economies in transportation......................................................................................... 19
4.
CROSS-COMMODITY VARIATION: TRANSPORT INTENSITY........................................ 21 4.1. The weight-value ratio .......................................................................................................... 21 4.2. Timeliness ............................................................................................................................. 22
5.
CONCLUSIONS .......................................................................................................................... 24
NOTES ................................................................................................................................................. 25 BIBLIOGRAPHY ................................................................................................................................ 26
West Lafayette, July 2004
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRANSPORTATION COSTS AND TRADE OVER TIME
- 11
1. INTRODUCTION
International economists have recently begun to focus on the role played by transportation costs in determining patterns of international specialisation and trade. There are large and growing literatures that address the impact of transportation on economic geography, the expansion strategies of multinational firms, the wisdom of regional versus multilateral trade liberalisation, and the distribution of bilateral trade. These questions are fundamentally cross-sectional in nature, asking how variation in costs at a point in time influences specialisation and trade. A slightly different question, and the focus of this paper, asks: how have these costs evolved over time? To address this question systematically, we begin by thinking about transportation costs as a wedge between prices inclusive and exclusive of shipping costs. This paper focuses on two kinds of relative prices, examines what we can learn from them and how they have evolved over time. The import price wedge compares the price of goods at the exporter’s departure port to the price at the importer’s point of delivery. That is, it describes how much higher are the prices of imports as a consequence of having to pay shipping costs. Alternatively, one can think of this wedge as a protective barrier behind which less efficient domestic firms hide from the encroach of foreign competition. This wedge helps determine how much a country imports, and the size of its gains from trade. Chapter 2 of this paper describes how the level of transportation costs, and therefore the size of the import price wedge, have evolved. The sourcing wedge compares two different foreign sources of supply: for example, the price of steel originating in Russia relative to steel originating in Korea when sold in a third market such as France. This wedge determines from whom an importer buys (or to whom an exporter sells). It may be driven by policy, such as preferential trading agreements, but also differences across sources in transportation costs. Chapter 3 of this paper describes the role of distance and transportation scale economies in determining the size of the sourcing wedge and its evolution. The evolution of both import and sourcing wedges depend on obvious factors such as technological developments and cost shocks, but also on the transportation intensity of the goods being moved. If we compare the cost of shipping $1 of coal to $1 of computer microchips, the transportation intensity of these cargos differs for two reasons. First, $1 of coal is bulkier and heavier than $1 of computer microchips, requiring greater stowage space and fuel expenditure to move. Second, microchips may require higher quality transportation services, greater care in handling and more rapid delivery than coal. To understand the evolution of import and sourcing wedges it is then necessary to understand how the composition of traded goods, and the intensity with which they use transportation services, has changed. Chapter 4 focuses on the evolution of the weight of trade and the demand for timeliness. It should be noted that no single source of data provides a definitive picture of the costs of transport. Only a small number of countries track transportation costs as part of their trade data, and only the United States has done so for a significantly long time span.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
12 – TRANSPORTATION COSTS AND TRADE OVER TIME
Lacking a single, comprehensive data set, this paper gathers together an eclectic mix of data on prices for international ocean and air transportation. Primary sources include: index numbers for ocean shipping prices, gathered from shipping trade journals; air freight prices constructed from survey data on air cargo; and freight expenditures on imports, collected by customs agencies in the United States, New Zealand and a number of Latin American countries. A major weakness of the study is that data on overland transport are not provided. These data are much harder to obtain and, in any case, regulation of and infrastructure for overland transport are sufficiently different across countries that information gleaned from one source would have little relevance to other markets.
2. THE IMPORT WEDGE: PRICES AT HOME AND ABROAD
In this chapter, we address the evolution in the import wedge — the price of goods at the exporter’s departure port relative to the importing destination. Among other things, understanding the evolution of this price wedge can be useful for understanding the role of declining transportation costs in explaining growing world trade. Economic historians have shown that substantial reductions in shipping costs played a key role in the first era of globalisation from 1850-1913. The post-war era has also been characterised by rapid trade growth. While the precise causes of post-war trade growth are not well understood, declines in transport costs top the list of usual suspects. We begin by examining the size of ad valorem transportation costs for several countries. We emphasize ad valorem rather than per unit costs as this is the relevant number for understanding the size of the import wedge. That is, we can write the value of imports valued cif (inclusive of shipping costs) at the point of delivery as p * q = ( p + f ) q , where q is the quantity shipped, p is the origin price, f is the shipping bill per quantity, and p* is the delivery price inclusive of shipping costs. Rearranging, the shipping cost gives the wedge, or difference, between the origin and destination prices, p * / p = 1 + f / p . In addition to describing the degree to which prices are higher for consumers of imports, the wedge describes how much “protection” domestic firms have from foreign competition. The data we examine come from customs declarations forms, in which the importing country requires the shipping firm to report the value of the shipment measured fob (free on board, or exclusive of shipping costs, or pq above) and cif. To simplify the reporting, we aggregate over multiple shipments and exporters to calculate the ad valorem freight bill, f/p, as the total freight bill paid, divided by total value shipped. Equivalently, this can be thought of as a trade-weighted average of shipping costs for each individual shipment. Table 1 reports ad valorem shipping costs — i.e. shipping costs expressed as a percentage of the value of goods shipped — for a number of countries in 1994 and 2000, in aggregate and by 1-digit SITC goods classification. Shipping costs create a substantial wedge between home and foreign prices for all countries except the US (this is primarily because US imports are dominated by North American goods with very low shipping rates). Small and land-locked countries pay higher costs than large countries with ocean access. Rates differ substantially across products, with ad valorem costs being much higher for bulk commodities than for manufactures.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
USA 3.3 7.7 5.2 7.5 4.1 6.6 3.0 5.1 1.9 4.9 1.0
Chile 15.5 12.5 7.0 13.2 28.9 12.1 8.8 9.8 6.8 6.5 6.4
Argentina 7.5 9.9 11.3 15.2 14.7 10.8 7.6 9.4 5.6 9.3 4.5
Argentina Brazil 8.3 10.6 9.4 9.2 7.1 5.5 12.3 7.2 29.7 15.7 7.8 6.2 6.4 5.3 7.1 6.8 4.2 4.1 6.4 5.7 11.4 8.7
USA 3.8 8.2 6.9 8.2 6.6 7.1 4.5 5.3 2.0 4.7 1.0 2000
Chile 8.8 12.7 8.4 12.0 11.8 9.3 10.2 10.9 6.3 9.1 7.6
Paraguay Uruguay 9.6 7.0 10.8 8.0 6.3 4.9 16.5 13.8 23.9 12.3 6.7 5.1 9.0 5.0 8.5 6.2 9.4 4.9 16.0 7.0 5.2
Brazil 7.3 10.4 9.0 7.7 10.7 5.4 6.8 8.5 5.1 8.1 0.8
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
Peru 13.4 13.5 5.5 12.8 23.4 16.0 9.0 9.9 6.4 7.0 6.4
N. Zealand 8.3 14.5 9.4 16.3 9.9 10.6 9.0 10.0 6.3 6.6 0.6
Mexico 3.6 6.0 7.3 5.5 6.9 7.3 6.6 13.2 7.6 19.5 1.7
Uruguay 4.6 3.6 4.8 3.7 4.7 2.6 3.0 4.7 4.1 5.8 2.5
Bolivia Colombia Ecuador 8.4 8.1 19.2 16.1 12.9 13.8 7.5 6.3 8.5 10.6 11.7 14.2 16.2 22.7 25.9 8.4 11.8 11.4 9.2 7.2 9.2 9.7 7.8 10.5 5.4 4.1 5.8 7.9 6.8 9.0 7.9 12.0 49.7
Paraguay 13.3 12.0 10.4 10.2 20.9 12.5 10.4 11.2 13.8 15.2 6.8
Note: Table reports shipping charges paid as a percentage of goods value shipped, aggregating over all exporters for each importer and commodity. Sources: US Census, Statistics New Zealand, ALADI Secretariat, ECLAC 811 Database.
All goods Food and live animals Beverages & tobacco Crude materials Mineral fuels, lubricants Animal & veg. oils, fats Chemicals & related prod. Manuf. goods (by material) Machinery & transp. equip. Miscellaneous manufactures All other goods, NES
All goods Food and live animals Beverages & tobacco Crude materials Mineral fuels, lubricants Animal & vegetable oils, fats Chemicals & related products Manufactured goods (by material) Machinery & transp. equipment Miscellaneous manufactures All other goods, NES
1994
Table 1. Ad valorem shipping costs : 1994 and 2000
TRANSPORTATION COSTS AND TRADE OVER TIME
- 13
14 – TRANSPORTATION COSTS AND TRADE OVER TIME
Total ad valorem costs are actually larger than these data suggest. These customs data typically cover only the international leg of transport, omitting inland charges. For shipments to and from coastal locations this will be accurate, but shipments inland will be more expensive. Case study evidence shows that international ocean freight comprises only a third of total door-to-door shipping charges, and this fraction has changed little over time. Also, the Table 1 numbers are trade-weighted averages of shipping costs. When importers choose export sources so as to minimise freight costs, the Table 1 numbers attach a large weight to unusually small costs and small weights to unusually large costs; i.e. the trade-weighted average understates true rates. Simple average freight rates for these countries are roughly double those reported in the table. How have these numbers changed over time? Figure 1 reports time series variation in aggregate ad valorem freight rates for the US and New Zealand. (A broader set of countries would be highly desirable but these are the only countries whose public use trade data includes a lengthy time series on shipping costs.) The longer New Zealand series show rates fluctuating between 7 and 11 per cent of import value but not declining over time. Data on freight charges in US trade show declines, but much of this is because the data series begins at the same time that the first oil shock dramatically increased shipping costs. While US data exhibit declining rates after 1974-98, imputing 1973 values from other sources eliminates nearly this entire decline.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRANSPORTATION COSTS AND TRADE OVER TIME
- 15
2.1. Modal prices These aggregate series fail to reflect a set of important changes in the modal composition of trade. In the US, airborne trade with partners outside of North America rose from 10 per cent of all shipments by value in 1965 to half of all shipments by 2002. Worldwide, the weight of trade that is air shipped has grown even faster. From 1963-96, seaborne trade in non-bulk products (measured in tonne-km performed) increased four-fold. In that same period, airborne trade (measured in tonne-km performed) increased 45-fold. Next, we examine changes in the price of air versus ocean shipping. 2.1.1
The cost of ocean shipping
We begin by providing index numbers for unit prices ($/quantity) of ocean-borne shipping. Price indices for ocean shipping are available from several sources, with varying coverage of time periods, goods shipped and routes. The indices reported here are chosen because they offer the longest time series of data. Two of the indices, constructed by the Norwegian Shipping News (NSN), cover voyage charter and time charter tramp shipping. A voyage charter is a contract to ship a large quantity of a dry bulk commodity between specific ports. The NSN voyage charter price index represents a weighted bundle of spot market prices ($/ton) for shipping major bulk commodities on several important routes worldwide. A time charter is a contract to employ the services of an entire ship for a set period of time (usually up to a year). The NSN time charter index reports a weighted bundle of spot charter prices for ships of various sizes ($/tonnage) in many ports worldwide. To evaluate the real costs of shipping over time, we deflate these indices using the US GDP deflator.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
16 – TRANSPORTATION COSTS AND TRADE OVER TIME
Figure 2 displays time series plots for time and voyage charters. Leaving aside very large price spikes in the oil shock years, and in the 1954-57 period, the time charter series shows no clear decline, while the voyage charter series exhibits downward trends in prices relative to the US GDP deflator. This suggests that the real price of bulk shipping, measured in dollars per tonne, has declined over time. However, when deflated by the price of the bulk commodities typically shipped via tramps, voyage charter rates are roughly constant and time charter rates are increasing. That is, the price of bulk commodities has fallen faster than the unit cost of tramp shipping, yielding no change or even increases in the ad valorem barrier to trade posed by international transport. While bulk commodities represent a large fraction of world trade by weight, they are a small and declining fraction of world trade by value. A third index, calculated by the German Ministry of Transport, measures liner shipping prices. The index heavily emphasizes general cargo, including containerised shipping and manufactured merchandise of all sorts, and so is more representative of the commodity composition of the majority of world trade. Figure 3 reports movement in the liner shipping price index relative to the German and US GDP deflator. Beginning in 1954, prices rise slightly through 1970, then sharply until a 1985 peak, with declines thereafter. The largest single rise occurs between 1973 and 1974. Note that rates rising through 1985 and declining thereafter closely match the aggregate New Zealand freight expenditure data from Figure 11. Though they rise and fall sharply, liner prices in 2000 remain very close to prices in 1970.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRANSPORTATION COSTS AND TRADE OVER TIME
2.1.2
- 17
Costs of air shipping
Annualised growth rates for world air freight revenues (measured relative to tonne-kilometres shipped), constructed from World Air Transport Statistics (WATS), are reported in Figure 42. The numbers show rapid declines in average revenues in the 1950s, 60s and 80s, with slow declines in the 90s and increases in the late 1970s.
The International Civil Aviation Organization (ICAO) “Survey of International Air Transport Fares and Rates”, published annually from 1973-93, contains rich overviews of air cargo freight rates for air travel markets around the world. In Table 2, this author reports annualised rates of change in air freight rates for each route group between 1973 and 1993. The nominal values from the ICAO Survey are deflated using the US GDP deflator to get a real price per kilogramme shipped, and by an index of air-shipped traded goods prices. The latter allows a calculation of changes in the ad valorem freight rates for air shipped goods. Measured in prices per kg shipped, worldwide we see an annual rate of decline of around 1.5 per cent. Prices decline on almost all routes, but they decline more on long routes and on those involving more developed nations. Declines in ad valorem rates are much larger than price per kg freight rates: 3.5 per cent per year over all routes. This is explained by substantial real increases in the price per kg of predominantly air-shipped goods.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
18 – TRANSPORTATION COSTS AND TRADE OVER TIME
Table 2. Changing air fares by region (annual rates of change) Shipping price per kg (tsso$)
Shipping price per $ shipped
1973-93
1973-93
All Routes
-1.53
-3.48
Developed Nation Routes: North Atlantic Mid Atlantic S Atlantic North and Mid Pacific South Pacific
-2.22 -1.26 -1.13 -2.39 -1.74
-4.16 -3.22 -3.06 -4.33 -3.69
Developing Nation Routes: North to Central America North and Central America to South America Europe to Middle East Europe and ME to Africa Europe/ME/Africa to Asia/Pacific
1.04 -0.14 -0.58 -1.13 -0.92
-0.97 -3.12 -2.56 -3.09 -2.88
Local Routes: Local Asia/Pacific Local North America Local Europe Local Central America Local South America Local Middle East Local Africa
-0.95 -0.80 -0.42 2.10 -0.83 -0.52 -0.14
-2.87 -2.77 -2.39 1.43 -2.80 -2.50 -2.12
Notes: (1) All series expressed in terms of annualised growth rates. (2) Price per kg and ad valorem freight rate series constructed using mean shipping distance within that route group. (3) Price per kg deflated using US GDP deflator. Ad-valorem rates constructed using a price per kg import price index.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRANSPORTATION COSTS AND TRADE OVER TIME
- 19
3. THE SOURCING WEDGE
In this chapter we address the sourcing wedge, which compares prices for two different foreign sources of supply. This wedge determines from whom an importer buys (or to whom an exporter sells). While a great many factors affect the size of this wedge, we focus on the role of distance and transportation scale economies in determining the size of the sourcing wedge and its evolution. 3.1. The role of distance Distance impedes trade to a surprising extent. Roughly half of world trade takes place between countries located within 3 000 kilometres of each other. Regression analysis affirms that the tradedistance relationship is robust to the inclusion of a wide variety of partial correlates, with typical estimates suggesting that doubling distances halves trade. Transportation costs are an obvious starting point for explaining this fact. It is clearly less expensive to ship from the UK to France than from Australia to France. Each km travelled requires greater fuel, manning and capital expenses. The question is, how much? The simplest way to calculate this is to estimate an elasticity of shipping costs with respect to distance, and to see how it has evolved over time. Hummels (1999) provides these estimates for air and ocean shipping at each point in time from 1974-98. Comparing a long (9 000 km) route to a short (1 000 km) route for comparable ocean-shipped commodities, the longer route was 59 per cent more expensive in 1974, but only 32 per cent more expensive in 1998. For air shipping, the longer route was three times as expensive in 1974, but only 68 per cent more expensive in 1998. In sum, costs are substantially rising in distance, but this effect has diminished over time. Further, the effect is much larger for air shipping than it is for ocean shipping. This is relevant because air shipping is more likely to be used the longer the route. The effect is a significant flattening in the relationship between shipping costs and distance. However, it should be noted that despite this change in shipping costs, the grip of geography on trade flows themselves remains. Distance affects trade as much, or more, today as it did forty years ago. The reasons for this are not entirely clear, but one intriguing possibility related to transportation intensity is discussed in chapter 4. In particular, the dollar cost of an extra km travelled may be falling, but the time cost may be rising as traded goods become more time sensitive. 3.2. Scale economies in transportation A second major factor in the size of the sourcing wedge is scale economies in shipping. The circumstantial evidence for investigating scale economies in shipping can be seen by examining freight costs for large versus small countries. Examining Table 1, we see that larger importers have smaller shipping costs for comparable goods. Or, consider Japan and the Ivory Coast, equidistant from the US West and East Coasts, respectively. US importers from the Ivory Coast pay shipping TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
20 – TRANSPORTATION COSTS AND TRADE OVER TIME
costs which are twice as high as those from Japan, even after adjusting for differences in the commodity composition of trade. More systematically, Hummels and Skiba (2004) use data from many importer-exporter pairs to estimate that doubling trade quantities results in a 12 per cent reduction in shipping costs. What is the source of scale economies in shipping? One possible source lies in the domestic trade infrastructure built up by each country. Ports (and the internal road or rail system necessary to reach them) tend to be large lumpy investments. If the fixed costs are large enough, an increased trade scale will benefit the investing country directly, and perhaps some of its trade partners, through lowered shipping costs. Scale economies may also operate at the level of the country pair and the trade route. The capacity of a modern ocean-going liner vessel is large relative to the quantities shipped by most exporters. As a consequence, goods are almost never shipped from point to point directly between the exporter and importer. Instead, a liner vessel may stop in a dozen ports in many different countries. As trade quantities increase it is possible to more effectively realise gains from several sources. First, a densely traded route allows for effective use of hub-and-spoke shipping economies – small container vessels move quantities into a hub where containers are aggregated into much larger and faster containerships for longer hauls. Examples include the European hub of Rotterdam, as well as Asian hubs in Singapore and Hong Kong. Second, the movement of some goods requires specialised vessels. Examples include ships specialised to move bulk commodities, petroleum products, refrigerated produce and automobiles. Increased quantities allow introduction of these specialised ships along a route. Similarly, larger ships will be introduced on heavily traded routes, and these ships enjoy substantial cost savings relative to older, smaller models still in use. (One source of scale advantage is in crew costs, which are roughly independent of ship size.) An historical example of these effects in combination can be seen in the introduction of containerised shipping. Containerised shipping is thought by many specialists to be one of the most important transportation revolutions in the 20th century. The use of standardised containers provides cost savings by allowing goods to be packed once and moved over long distances via a variety of transport modes (truck, rail, ocean liner, rail, then truck again) without being unpacked and repacked. Despite these advantages, containerised shipping did not diffuse immediately throughout the world. Instead, it was first introduced in the US in the 1960s, then on US-Europe and US-Japan routes in the late 60s and 70s, then to developing countries from the late 70s onward. The degree of containerisation varies markedly across regions, and is positively correlated with route GDP. An obvious explanation for this slow diffusion lies in the fixed costs of adoption. To make full use of containerisation requires container-ready ocean liners and ports adapted to container use (specialised cranes, storage areas and rail-heads). Building container ports typically requires large capital expenses, and will not be undertaken unless a large volume of trade can be moved through them. Similarly, shipping companies will not dedicate a container-ready ocean liner to a route unless there is a sufficient volume of trade along that route. Finally, the full benefit of containerisation may not be enjoyed unless it is combined into hub-and-spoke systems at the regional level that allow the matching of differently sized ships to appropriate route lengths. Thus, adoption of a revolutionary shipping technology like containerisation depends on the scale of trade at the level of the exporter, and the exporting route.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRANSPORTATION COSTS AND TRADE OVER TIME
- 21
Apart from the advantage it imparts to large countries, the existence of scale economies in transport has several interesting implications. While trade growth can lead to higher transport prices, especially as port facilities suffer congestion, it can also lead to scale benefits as described here. This suggests that trade liberalisation can lead to a virtuous cycle: tariff reductions spur trade growth, which lowers transport costs, leading to further growth. Also, to the extent that shipping scale economies are regional in scope, transportation facilities become a kind of regional public good. That is, increases in trade between the US and Argentina benefit these partners, but also the other countries (Brazil, Venezuela, the Caribbean) lying along the route. It follows from this that countries may prefer to see tariff liberalisation concentrated among regional neighbours.
4. CROSS-COMMODITY VARIATION: TRANSPORT INTENSITY
Cost shocks and technological innovation play an important role in the time series changes in transportation costs. Another critical factor is changes in the transport intensity of traded goods. In this chapter we discuss two significant changes in transport intensity. The first relates to the value of transportation services needed to move $1’s worth of a good. The second relates to changes in the quality of transportation services, in particular, delivery times. 4.1. The weight-value ratio Transportation specialists are accustomed to thinking of transportation costs in per-unit terms: the cost of transportation services necessary to move grain a tonne-km or to move one TEU container from Rotterdam to Hong Kong. International trade specialists who pay attention to shipping costs as an impediment to trade, are accustomed to thinking of these costs in ad valorem terms: the cost of transportation services necessary to move a dollar of grain or microchips between two points. The distinction is important because even if the cost of moving one TEU remains constant, the ad valorem cost and the implied impediment to trade can change if the contents of the container change. To see this, suppose we sell q TEU containers of a good at a price p, and pay shipping costs f per container shipped3. What is the ratio of destination to origin prices?
p * / p = ( p + f )q / pq = 1 + f / p If the container holds scrap metal, p is low and the ratio p*/p is high. If the container holds micro chips, p is very high and the ratio p*/p is close to 1. This observation is important because the commodity composition of world trade has shifted dramatically in the last thirty years. Trade in manufactures has grown much more rapidly than trade in bulk commodities. As this happens, the weight/value ratio of trade drops, and with it, one measure of transport intensity drops as well. Between 1970 and 1999, WTO data show that the real value of trade in all products grew 18-fold in real terms, while trade in manufactures grew 22-fold, and trade in agricultural and mining goods TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
22 – TRANSPORTATION COSTS AND TRADE OVER TIME
grew ten-fold. Meanwhile, tonnage moved via seaborne and airborne shipping together slightly more than doubled4. Altogether, that implies that the weight/value ratio of total trade fell by a factor of nine since 1970. One implication of the drop in weight/value ratio is the shift toward air shipping and away from ocean shipping. Air shipment is more likely to be used when the ad valorem price differential between the two modes is small. That is most commonly the case when the freight bill is a small fraction of the value of the good, that is, when the weight/value ratio for the product is small. Consider this example. A consumer wants to import a $16 bottle of wine from France. Air shipping costs of $8 are twice ocean shipping costs of $4. Going from ocean to air increases the delivered cost by $4 or 25 per cent. Now suppose the consumer’s tastes improve and he wants to import a $160 bottle of wine from France. The shipping costs are the same, but now the $4 cost to upgrade to air shipping represents just a 2.5 per cent increase in the delivered price. The consumer is much more likely to use the more expensive shipping option when the effect on delivered price is smaller. The broader point is that transportation demand is derived from import and export demand. No-one values transportation directly, they value it only as part of a process of getting internationally traded goods to their final consumers. And those consumers are sensitive to changes in the delivered price, not to changes in the transportation price. If the cost of transportation substantially affects the delivered price, as in the first example, modal choice will be driven by cost considerations. But if the transportation price is but a small fraction of the delivered price, it will likely be trumped by other factors such as timeliness or reliability. The same lesson is true of all cost differentials related to transportation. When shippers are deciding which modes to use, or which ports to use, they look at cost differentials. But if these cost differentials have a minor effect on the delivered price, other factors will get greater weight. The wine example just given also illustrates a second implication of the weight-value ratio for trade. Consider the size of the sourcing wedge, the relative price of goods from two sources when shipped to a third destination, when those goods are of different quality and price. France pays $10 to ship a $50 bottle of wine to the US, while Chile pays $10 to ship a $10 bottle. Exclusive of shipping costs, the French wine is 5 times more expensive. Inclusive of shipping costs, the French wine is only 3 times more expensive ($60/20). The higher the shipping costs, the smaller is the gap between the delivered prices of high- and low-quality goods. Hummels and Skiba (2004) show evidence that this effect plays an important role in two places. As in the French/Chile wine example, high shipping costs cause importers to shift demand to high-quality sources. High shipping costs also cause a particular exporter to ship high-quality wine to distant destinations while reserving lower-quality wine for the domestic market. As shipping costs rise or fall over time, these quality shifting effects rise and fall with them. 4.2. Timeliness Up to this point we have focused on the cost of shipping a good, measured either in quantity or value terms, taking as fixed the quality of the transportation service. However, there have been pronounced changes in the quality of international transport over the past thirty years, the most notable being transportation time. Today, shipping containers from European ports to the US Midwest requires 2-3 weeks; to Far Eastern ports as long as six weeks. In contrast, air shipping requires only a day or less to most destinations. Ocean liner service itself has become much faster than in years past, both because the ships are larger and faster, and because their loading and unloading times are dramatically lowered by containerisation.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRANSPORTATION COSTS AND TRADE OVER TIME
- 23
How valuable is timeliness? Two recent empirical papers shed light on this point. Evans and Harrigan (2003) show that timeliness in the apparel industry has a pronounced effect on sourcing patterns. Clothing lines that have high restocking rates within a given buying season are much more likely to be sourced locally than those in which orders are taken once per season. Hummels (2001) estimates a demand for timeliness by examining the premium that shippers are willing to pay for speedy air shipping relative to slow ocean shipping. He shows two effects. First, for every day in ocean travel time that a country is distant from the importer, the probability of sourcing manufactured goods from that country drops by 1 per cent. Second, conditional on exporting manufactures, firms are willing to pay just under 1 per cent of the value of the good per day to avoid the travel delays associated with ocean shipping. Why are these effects so large? The per-day time cost of the good is a function of two factors. The first is the per-day interest rate on the good in transit, otherwise known as pipeline inventory. The second factor is a “depreciation rate” for the good. The depreciation rate encompasses any reason that a newly-produced good might be preferable to an older good. Obvious examples include spoilage that is literal and predictable, such as with fresh produce or cut flowers. Depreciation may also reflect the immediate need for the good, and lost profitability/utility from the good if it is not available. More generally, with long lags between production ordering and final sales, firms may face a mismatch between what consumers want and what the firm has available to sell. Consumers will pay a premium to purchase goods containing “ideal” characteristics, but firms may not be able to predict far in advance what constitutes the ideal. Firms that can wait longer to produce are better able to match the ideal characteristics and capture that premium. This can be accomplished either by producing locally, or by producing at a distance and air shipping. Specific examples of goods with this property may be useful here. Toy manufacturers generally do not know in advance which toys will emerge from among hundreds of competitors to capture the hearts and minds of children during the holiday gift-giving season. The “ideal” types command price premia over the non-ideal types. As firms near the holidays, they receive market signals (product reviews, early sales) about the ideal type, and can adjust accordingly. As Evans and Harrigan (2003) demonstrate, apparel is another example where ideal characteristics are difficult to discern well in advance, and firms must produce (and ship) much closer to sales dates, or restock mid-season. Finally, personal computers exhibit extreme time sensitivity of this sort. Standardised packages do not appeal to many consumers, who are willing to pay more for a customised computer, manufactured to particular specifications (CPU speed, screen size, amount of RAM). So manufacturers simply do not build the computer until they know the precise ideal characteristics. There are several factors that may explain increases in time sensitivity in the past decades. The first is the commodity composition of trade. The estimates in Hummels (2001) show that bulk commodities and simple manufactures exhibit no time sensitivity. As trade shifts from these products toward more complex manufactures, time sensitivity grows. Trade has grown most rapidly in those products with the highest estimated time sensitivity. (Note, however, that the causality may go the other way. Trade in time-sensitive goods may be rising precisely because the cost of rapidly shipping them has fallen so dramatically.) Second, time sensitivity is rising even within particular goods. As consumers grow richer, their willingness to pay for precise product characteristics grows. That in turn puts pressure on manufacturers to produce to those specifications, and be rapidly adaptable; and for richer consumers fast delivery times become part of the expected service quality. TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
24 – TRANSPORTATION COSTS AND TRADE OVER TIME
Finally, the post-war era has seen rapid growth in other forms of integration, in particular in foreign direct investment and vertical specialisation/fragmentation. Vertical specialisation, or more broadly multi-stage production, may be especially sensitive to lags and variability in timely delivery. The absence of key components can render idle an entire assembly plant. An emergency shipment that arrives in a timely fashion may be worth many times the nominal price of the component, while late arrivals are of considerably depreciated value. Inventory on-hand must also be increased in order to accommodate arrival time variations. This in turns magnifies the costs of defects in component quality, as sizeable inventories (at the plant, in transit) may be built up before defects are detected. The defect problem motivates “just-in-time” inventory techniques, which aim to minimise both the inventory on hand and in the pipeline. Clearly, the ability to implement a just-in-time strategy is limited when parts suppliers are a month of ocean transit time removed from the assembly plant. The broad point is that rapid declines in air transport costs, and the corresponding reduction in the cost of time-savings, may be responsible for the growth of time- and coordination-intensive forms of integration. And the growth in time- and coordination-intensive forms of integration puts further demands on timeliness in delivery.
5. CONCLUSIONS
In this paper, we have discussed the evolution of international transportation costs. The major findings are these: •
Transportation costs create a significant wedge between prices of domestic and foreign goods, and between prices of different foreign goods in home markets. The size of this wedge differs greatly across countries, depending on their distance to markets, level of development, commodity composition of trade and market size.
•
In aggregate, these costs have declined over time, owing principally to very rapid declines in the price of air shipping, and a compositional shift in trade toward light manufactures.
•
However, while trade has become lighter, and therefore less transport-intensive per dollar shipped, it has also become more time sensitive.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRANSPORTATION COSTS AND TRADE OVER TIME
- 25
NOTES
1.
Anecdotal and case study evidence suggests that the especially rapid liner price increases that occurred in the 1970s for German trade appear to have occurred more broadly. UNCTAD’s annual Review of Maritime Transport reported annual liner rate increases of 10-15 per cent across nearly all routes. Studies of rates on US North Atlantic liner routes found real increases (relative to the dollar deflator) ranging from 21 to 26 per cent between 1971 and 1975, comparable to those found in German trade. Sletmo and Williams (1981) report nominal rate increases of between 61 and 103.5 per cent.
2.
The latter calculation assumes a cost technology for air shipping given by C=at(ton)(km).5 where at is a time-specific cost shifter. If the elasticity of costs with respect to distance shipped is less than one, doubling the distance shipped results in a decline in average costs per tonne-miles. WATS data indicates a rapid rise in mean distance shipped over time, and so average revenue must be adjusted accordingly.
3.
Of course, if shipping costs are proportional to the price shipped, then the ratio p*/p is the same with or without shipping costs. Hummels and Skiba (2004) estimate an elasticity of freight prices, with respect to the price of goods shipped, of 0.14. This means that a rise in the price of the good creates only a small increase in shipping prices.
4.
If we consider only non-bulk dry cargos, the weight of seaborne and airborne cargos together grew by a factor of 3.5. This calculation ignores growth in overland (rail, road) international shipping. If land-based shipping represented a large and growing share of world trade, this calculation would overstate the drop in the weight/value ratio. However, for most countries, overland trade is either unimportant in terms of its weight share of trade or, for countries like Germany, the weight share of overland trade has been roughly constant over time.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
26 – TRANSPORTATION COSTS AND TRADE OVER TIME
BIBLIOGRAPHY
Evans, Carolyn and James Harrigan (2003), “Distance, Time, and Specialization”, NBER No. 9729. Hummels, David and Alexandre Skiba (2004), “Shipping the Good Apples Out? An Empirical Confirmation of the Alchian-Allen Conjecture”, Journal of Political Economy, Vol. 112, pp. 1384-1402, University of Chicago, December. Hummels, David (2001), “Time as a Trade Barrier”, mimeo, Purdue University. Hummels, David (1999), “Have International Transportation Costs Declined?”, mimeo, University of Chicago.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES -
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES
John S. WILSON Xubei LUO Harry G. BROADMAN The World Bank Washington, DC United States
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
27
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES -
29
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES
SUMMARY
ABSTRACT .......................................................................................................................................... 31 1.
INTRODUCTION ......................................................................................................................... 32
2.
TRADE AND REGIONAL INTEGRATION IN EUROPE: EMPIRICAL ANALYSIS FOR CAPACITY BUILDING ............................................................................................................... 33 2.1. Description of the database .................................................................................................... 34 2.2. Cross-country comparisons in ECA....................................................................................... 34
3.
BENCHMARKING THE RELATIONSHIP BETWEEN ECONOMIC DEVELOPMENT AND TRADE FACILITATION ................................................................................................... 37 3.1. Does economic growth enable building trade facilitation capacity?...................................... 37 3.2. Does trade facilitation promote economic growth? ............................................................... 40
4.
EXTERNALITIES FROM INVESTMENT IN TRADE FACILITATION ................................. 42
5.
CONCLUDING REMARKS – INFORMING CAPACITY-BUILDING POLICY PRIORITIES.................................................................................................................................. 49
ANNEXES ............................................................................................................................................ 51 NOTES .................................................................................................................................................. 55 BIBLIOGRAPHY ................................................................................................................................. 56
Washington, October 2004
The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. The analysis in this paper will be referenced in a forthcoming book by the World Bank, entitled: Disintegration and Re-Integration of Eastern Europe and Central Asia in the World Marketplace. TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES -
31
ABSTRACT
This paper examines the impact of trade facilitation on bilateral trade flows, building on the analytical framework and results in Wilson, Mann and Otsuki (2004). We examine trade facilitation and capacity building priorities in twelve countries in the Europe and Central Asia region – eight of the current members of the European Union: Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia and Slovenia, and three candidate members: Bulgaria, Romania and Turkey. Our results suggest that behind-the-border factors play an important role in determining bilateral trade flows (controlling for the effects of tariffs, development levels, distance and regional characteristics of exporters and importers, among other factors). The development of new data sets to expand work related to trade facilitation, including strengthening the empirical work explored here, is a key priority without which intelligent policy and priorities cannot be made. Our analysis is based on data from the World Economic Forum, Global Competitiveness Report 2001-2002, the World Competitiveness Yearbook 2000 and Kaufmann, Kraay and Zoido-Lobaton (2002). The results outlined here indicate that more gains in exports than in imports are expected, should the values of three out of the four indicators (port efficiency, regulatory regimes and IT infrastructures) of the new and candidate member countries improve halfway to the EU-15 average. These countries would expect large trade gains as well as improvements in trade balances as their integration into the EU continues. For example, the greatest absolute trade gains – $49 billion and $62 billion respectively – could be expected if their port efficiency and IT infrastructure reach half-way the average level of EU, and 70 per cent of trade gains are associated with export expansion. The analysis further suggests that priorities for capacity building differ with the two sets of countries examined. Among the four trade facilitation indicators, IT infrastructure improvement will lead to the largest gain for the new member countries, as well as for candidate member countries. If all indicators improve to halfway of the EU-15 level, almost 40 per cent trade gains will result from the improvement of IT infrastructure. As to the trade gains that could result from improvement in the other trade facilitation activities examined, however, they differ between the new and candidate member countries. The new members of the European Union exhibit large potential gains to trade with investments in port efficiency (both air and maritime ports), which represent a third of the total trade gain. The results for the candidate members suggest more widely dispersed gains with investments in port efficiency, customs regimes and regulatory policy. Improvements in each dimension share around 20 per cent of the total trade gains. Development planning and capacity-building programmes can take these results into consideration – along with many other factors – as integration into the European Union continues. In general, improvements in port facilities and IT infrastructure are likely to be more costly than the administrative reforms at the centre of customs regimes and regulatory policy – but they can have correspondingly high payoffs. The eligibility for additional EU financing with accession should provide more scope for improvements in these areas. For the candidate member countries, such as Bulgaria and Romania, that are currently negotiating the formal accession with the EU, it may be of particular interest to explore acceleration of telecommunications liberalisation and investment, for example. TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
32 - TRADE AND TRANSPORT FACILITATION:
EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES
1. INTRODUCTION
Understanding the relationship between economic development and the transport-related costs of international trade is relatively straightforward in theory. Analysis of how, in practice, modern trade logistics influence the facilitation of international commerce, however, is more challenging in empirical design and estimation: the linkages between the two are multifaceted, subtle and complex. Economic development and poverty alleviation are both achieved through income growth. We also know – from an increasing body of empirical evidence – that economic growth expands with world trade. Lower transport and other trade-related transaction costs, in turn, provide the engine through which trade expands to achieve development goals. The traditional definition of “trade facilitation” centred on ways to achieve lower international transport costs. In modern commerce, however, a broader definition is called for. Today, facilitating trade involves not only improved efficiency in logistics at ports and customs – through greater transparency, ensuring operational decisions are rules-based (rather than discretionary) and the use of advances in technology (including but not limited to IT), among other things – but also streamlined regulatory policy, deeper harmonization of standards and conformance to international norms so that overall transaction costs are lowered (Woo and Wilson, 2000; Wilson et al., 2002). Indeed, reducing the “behind the border” barriers associated with achieving the goal of lowering transaction costs through domestic reforms is increasingly at the of international policy deliberations in governments of transition economies and developing countries, among donor agencies and in regional and multilateral trade negotiating forums. In the transition countries of Europe, the CIS and Central Asia – hereafter “ECA” – reducing such barriers is increasingly seen as the key policy priority to accelerate integration into the world economy, including through accession to the World Trade Organization (WTO) and membership in the European Union (EU). Thus, many ECA countries are faced with the increasingly broader challenge of facilitating trade through moving goods through ports more efficiently by streamlining the movement of documentation, enhancing the professionalism of customs officials, harmonizing product and technical standards with international or regional regulations and strengthening the integration of new technologies into the transport and communications infrastructure. Meeting these “domestic” international trade facilitation challenges places enormous importance on the need for capacity building. As countries – and the international donor community – decide on how best to deploy resources for such capacity-building, a critical policy question arises: what is the relative impact of improvements in trade facilitation in contrast to gains from lowering traditional trade barriers, such as tariffs and quotas? Trade facilitation was added to the policy dialogue on trade issues at the Singapore Ministerial of the WTO in 1996. In August 2004, the WTO decided to focus part of the negotiations currently underway in the Doha Round on trade facilitation issues. As a result, today, trade facilitation discussions are at the centre of the Doha Development Agenda. The European Union, among others, has been a leading advocate of such negotiations. At the same time, regulatory reform, modernisation of customs regimes and infrastructure investment related to lowering trade logistics costs are key components of TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES -
33
development strategies for the new members of the EU and those in line for accession. Clearly, then, earmarking sufficient resources – both financial and human – to build the requisite institutional capacity in order to implement trade facilitation reforms, is increasingly an important policy issue for governments. There are a number of important challenges in carrying out empirical research on the impact of trade facilitation initiatives to help inform policy decisions and capacity-building priorities. They include selecting the most appropriate measures to gauge the extent of trade facilitation, identifying a suitable modelling methodology to estimate the impact of trade facilitation on exports and imports and establishing an analytical framework to estimate the effect of improved trade facilitation on trade flows. This paper builds on Wilson, Mann and Otsuki (2004) [hereafter WMO (2004)] and uses four separate indicators to estimate trade gains due to trade facilitation progress in (i) port efficiency; (ii) customs regimes; (iii) regulatory policy; and (iv) information technology infrastructure for new member and candidate members of the European Union. We also provide policy suggestions on the priorities in trade facilitation improvement.
2. TRADE AND REGIONAL INTEGRATION IN EUROPE: EMPIRICAL ANALYSIS FOR CAPACITY BUILDING
The new members of the European Union – Cyprus, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Slovakia and Slovenia – now have more direct access to regional markets. Four candidate EU members – Bulgaria, Croatia, Romania and Turkey – are preparing the groundwork for accession. The reduction in tariffs, quotas and other traditional steps toward integration are only part of the measures that both groups of countries are introducing. Measures are also being put in place to enhance port efficiency, improve customs regimes, reform regulatory policies and develop the information technology infrastructure. Conventional gravity model analysis suggests that transactions costs impede the exchange and the transfer of goods and services from different countries/regions. The wedge between export and import prices reduces profit margins. In particular, trade barriers – both tangible and intangible – limit trade and slow prospects for regional development. According to Overman et al. (2001), access to foreign markets alone could explain some 35 per cent of the cross-country variation in per capita income. Regions with higher transactions costs exhibit slower growth (Diamond, 1997; Limao and Venables, 2001; Redding and Venables, 2003). Importantly, with advances in technology, transport costs have become less subject to distance. Hummels (1999) suggests that, while in 1974 shipping commodities over a distance of 9 000 km by sea was approximately 60 per cent more expensive than shipping over a distance of 1 000 km, the cost difference was reduced by a half (to 30 per cent) by 1998. Given that the new members and candidate members of the EU are relatively physically far from the central markets of the EU, this reduction in long-distance transport costs would tend to facilitate trade significantly, all other things being equal. This reduction in “effective distance” raises even further the relative importance of lowering the transaction costs of trade. TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
34 - TRADE AND TRANSPORT FACILITATION:
EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES
Integration into the EU will, over time, clearly engender enhanced trade facilitation. EU membership will make it possible to move goods freely between member states without the need to complete formal import and export documentation or pay import VAT or customs duties. This should result in a more rapid movement of goods (and services) between member states and reduce transaction costs. The harmonization and implementation of the acquis communautaire also require new member countries to make major improvements in their overall economic environment – both at the border and behind the border. On a global basis, the estimates by WMO (2004) suggest that improvements among the four dimensions of trade facilitation that are the focus of this paper – port efficiency, customs regimes, regulatory policy and information technology infrastructures – can lead to significant trade gains. Their analysis indicates that for the 75 sample countries, raising capacity halfway to the global average would yield a $377 billion gain to world trade. Our paper takes as the starting point the same analytical framework underpinning the global assessment, but focuses on a subset of twelve ECA countries – eight of the current members of the EU: Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia and Slovenia (hereafter EU8); and four candidate EU members: Bulgaria, Croatia, Romania and Turkey1. We also move beyond the evaluation of trade facilitation based on a single parameter, such as the price of imports, the productivity of the transport sector or the costs of transportation, and examine the four dimensions of trade facilitation noted above. We use a gravity model of bilateral trade flows for our estimations, rather than a computable general equilibrium (CGE) approach. The scenarios examined here do not assume that all countries in our sample (those that have acceded to the European Union or candidate members) improve capacity by the same amount. Some ECA countries have further to go to reach best practice in, for example, regulatory reform or port efficiency, than do others. Moreover, to keep the simulated scenarios more realistic, we also assume that the countries initially less developed in trade facilitation are able to achieve only a relatively low level of trade facilitation improvements. 2.1. Description of the database We utilise the database in WMO (2004)2. Our measures indicate how far a country's performance is from the best–practice country in each of the four trade facilitation dimensions. The best-practice country is indexed to a value of 1.0. Among the 75 sample countries in 2001, Singapore is the best performer in port efficiency and Finland is the best performer in the other three areas. For each country, the indicator "Port Efficiency" is an average of the efficiency of the port facilities/inland waterways and air transport3; the indicator "Customs Regimes" captures the hidden import barriers other than published tariffs and quotas, and the irregular extra payments or bribes connected with import and export permits; the indicator "Regulatory Policy" is constructed as the average of the transparency of government policy and the control of corruption; and the indicator "Information Technology Infrastructure" is the measure of the speed and cost of Internet access and the contribution of Internet to the reduction of the inventory cost. 2.2. Cross-country comparisons in ECA The 15 EU member countries (EU-15) – the members that existed prior to May 2004 – are relatively advanced in all four areas, with an average value of 0.82, 0.87, 0.79, and 0.78 in port efficiency, customs regimes, regulatory policy and information technology infrastructure, respectively4. The development level of trade facilitation in the EU-15 ranks well on average; with customs efficiency rated the highest. TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES -
35
The new EU-8 member countries, however, are less developed in these four areas, with an average value of 0.60, 0.73, 0.65 and 0.64 (development level of 72%, 83%, 81%, 81% of the EU-15 respectively, see Figure 1). As to three candidate members (Bulgaria, Romania and Turkey), the development of their trade facilitation is further behind, with the state of their customs regimes the furthest behind – 58 per cent level of the EU-15. The level of development of the new and candidate member countries varies most in port efficiency – Estonia and Latvia are the best new performers with a development level of around 70 per cent of the best performers. The Slovak Republic and Bulgaria are at a level of less than 50 per cent of the best performers. Regarding IT infrastructure, Estonia performs well (reaching the level of 80 per cent of the best performer), followed by the Czech Republic, Slovenia, Slovak Republic and Romania (around 65-70 per cent of the best performer). Estonia and Hungary have the highest benchmark in customs efficiency. As to regulatory policy, the development levels of the new and candidate member countries tend to be less heterogeneous. Figure 1. Comparison of new and candidate EU members and the EU-15
0,90 0,80 0,70 0,60 0,50 0,40 0,30 0,20 0,10 0,00 port efficiency
customs regimes
regulatory policy
Czech Republic
Estonia
Hungary
Latvia
Lithuania
Slovak Republic
Slovenia
Bulgaria
Romania
Turkey
IT infrastructure Poland
Figure 2 shows that the new member countries, for example, the three largest economies (the Czech Republic, Hungary and Poland), are not only less developed than the EU-15 in trade facilitation as a whole, but also constrained in particular dimensions. Hungary’s customs regime approaches 95 per cent of the EU-15 level, while in ports efficiency it is about 60 per cent of the EU-15 level. The Czech Republic is relatively developed in IT infrastructure, which reaches almost 90 per cent of the EU-15 level, while in port efficiency it is less than 70 per cent of the EU-15 level. Poland, the least developed among these three, exhibits a level in the four areas of around 70 per cent of the EU-15 benchmark. In sum, the results suggest that in order to achieve the trade facilitation levels of the EU-15, the new member countries face differing challenges.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
36 - TRADE AND TRANSPORT FACILITATION:
EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES
Figure 2. New member countries (Czech Republic, Hungary and Poland)
Port efficiency
1
0.5
IT infrastructure
0
Customs regimes
Regulatory policy
EU15
Czech Republic
Hungary
Poland
Figure 3 indicates a similar situation: the countries that are candidate EU members suffer from low trade facilitation development overall and individual specific constraints. Bulgaria does relatively well in regulatory policy, which reaches almost 85 per cent of the EU-15 level, while its port efficiency index is less than 50 per cent of the EU-15 level. Romania, which performs well in ports and in IT infrastructure, with development levels of over 80 per cent of the EU-15, lags behind in customs performance. Figure 3. Candidate members of the EU (Bulgaria, Romania and Turkey) Port efficiency 1
0.5
IT infrastructure
Customs regimes
0
Regulatory policy
EU15
Bulgaria
Romania
Turkey
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES -
37
3. BENCHMARKING THE RELATIONSHIP BETWEEN ECONOMIC DEVELOPMENT AND TRADE FACILITATION
The new members of the EU as a whole have exhibited rapid economic growth in the last several years, despite the relatively weak performance of Western Europe and the world economy. Among the EU-8, real GDP growth was approximately 2.5 per cent in 2001 and 2002, and over 3 per cent in 2003 (World Bank, 2004). Along with income convergence (though a large gap still exists), the EU-8 have made substantial progress in other fields, such as reductions in inflation. 3.1. Does economic growth enable building trade facilitation capacity? One assumes that economic growth and efficiency in trade facilitation are positively related. The more developed a country is, the more resources it can devote to investing in trade facilitation capacity. By the same token, it is likely that the larger the economy, the higher is the rate of return on investment in improving trade facilitation. This positive relationship between growth and trade facilitation indicators is confirmed by the data in the 75 sample countries – the correlations between GDP per capita (adjusted by PPP) and port efficiency, customs regimes, regulatory policy and information technology infrastructures are, respectively, 0.78, 0.86, 0.68 and 0.81. The new and candidate members of the EU are less developed than the EU-15. Does their gap in economic development account in some way for the lagged progress in their trade facilitation capacity? Considering GDP per capita, is the development of trade facilitation in the new and candidate members of the EU what would be expected? Our analysis of the data suggests that the new and candidate EU member countries are in general weak performers in all four of the trade facilitation dimensions examined (Figures 4a-d). Indeed, even in the context of their relatively low level of economic development, their trade facilitation development is relatively low, in fact under the benchmark level. The only exception is Estonia, which performs more strongly than the benchmark level in all four indicators. Compared with Hungary, a country of similar economic indicators, Estonia is 40 per cent more developed in port efficiency, 30 per cent more developed in IT infrastructure and 20 per cent more developed in regulatory policy5. In fact, the trade facilitation level of Estonia is even more developed than the average of EU-15 in IT infrastructure. The strong growth in the electronics and telecommunications sectors has likely contributed to the recent economic performance of Estonia – the real GDP growth rate was 4.8 per cent in 2003, far above the regional average level6. Figure 4a shows that, besides Estonia, three countries (Latvia, Romania and Turkey) perform above their benchmark levels in port efficiency. The other new and candidate EU member countries are lagging behind; in particular, the Slovak Republic and Hungary exhibit a lower degree of trade facilitation capacity, despite their relatively high economic development level. Figure 4b shows that Hungary, the Slovak Republic and Lithuania – as well as Estonia – perform above their benchmark level in customs regimes. The other countries cluster close to the benchmark level. Figures 4c and 4d show respectively similar phenomena to Figures 4b and 4a. In other words,
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
38 - TRADE AND TRANSPORT FACILITATION:
EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES
apart from Estonia, many new and candidate member countries perform weakly in port efficiency and in IT technology – after taking their economic development level into account. Figure 4a. Benchmarking to the value of GDP per capita (port efficiency)
Figure 4b. Benchmarking to the value of GDP per capita (customs regimes)
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES -
Figure 4c. Benchmarking to the value of GDP per capita (regulatory policy)
Figure 4d. Benchmarking with respect to the value of GDP per capita
(information technology infrastructure)
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
39
40 - TRADE AND TRANSPORT FACILITATION:
EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES
3.2. Does trade facilitation promote economic growth? Our analysis also suggests that barriers to trade facilitation in the new and candidate EU members weaken their development potential. For example, export growth is one of the most important factors that contributed to recent economic growth in the Czech Republic, Hungary and Poland – the three largest economies in the EU-8. If infrastructure is upgraded and transaction costs lowered, trade volumes can expand. Expanded trade can also contribute to the reduction of unemployment rates7. Benefits to growth are associated with both the reduction in transport costs and the increased reliability in delivery times. In a modern economy, the delivery of goods and services are highly time-sensitive, with just-in-time supplies playing an important role in enhancing productivity and private sector efficiency. WMO (2004) develops a gravity model to examine the role of trade facilitation in bilateral trade (see Annex 2 for a description of the model). Following that exercise, we simulate the impact of the hypothetical improvements in port efficiency, customs regimes, regulatory policy and information technology infrastructure half-way to the benchmark level on bilateral trade flows. Our results here suggest that behind-the-border factors do play a critical role in determining bilateral trade flows, after controlling the effects of tariffs, development level, distance and regional characteristics of exporters and importers. Others factors held constant, the improvement of port efficiency, customs regimes, regulatory policy and IT infrastructure will lead to the increase of both export and import volumes. The analysis indicates that the increase in a country’s export (import) volumes, engendered by higher levels of its behind-the-border indicators, depends on (i) the marginal impact of the indicator in question on exports (imports); (ii) the relative improvement in this indicator; and (iii) the initial export (import) volumes. Table 1 shows that the new and candidate EU member countries are expected to have large trade gains if they raise their trade facilitation capacities to a level that is 50 per cent of the EU-15 average. The lower the initial trade facilitation capability of a country, the larger the improvement it is expected to make in order to reach the benchmark level (i.e. half-way to the EU-15 average). Our results show that more gains in exports than in imports are expected, should the values of three out of the four indicators (port efficiency, regulatory regimes and IT infrastructures) of the new and candidate member countries improve to halfway to the EU-15 average. They would expect large trade gains as well as improvements in trade balances as their integration into the EU continues. For example, the greatest absolute trade gains – $49 billion and $62 billion, respectively – could be expected if their port efficiency and IT infrastructure reach half the average level of the EU, and 70 per cent of trade gains are associated with export expansion. Table 2 shows that improvements in behind-the-border factors also result in large relative trade gains. Such gains reach around 11 per cent if the capacity of the new and candidate EU members reach halfway to the EU-15 average. For instance, Lithuania’s trade volume rises more than 25 per cent if its IT infrastructure level reaches 50 per cent of the EU level.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES -
41
Table 1. Absolute trade gains for new and candidate EU members– Raising capacity half-way to the EU average ($ billion) Country
Port efficiency Export Import
New members: Czech Republic Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia Sub-total: Candidate members: Bulgaria Romania Turkey Sub total: TOTAL:
Customs regimes Export Import
Regulatory policy Export Import
It infrastructure Export Import
6.55 0.21 7.99 0.17 0.60 6.61 5.32 1.11 28.55
2.23 0.09 2.82 0.09 0.29 2.91 1.87 0.37 10.67
2.23 0.05 0.33 0.20 0.15 3.07 0.44 0.27 6.74
2.28 0.06 0.35 0.32 0.22 4.07 0.47 0.27 8.04
2.48 0.10 3.03 0.10 0.12 4.13 1.00 0.62 11.57
1.15 0.06 1.46 0.07 0.08 2.48 0.48 0.28 6.06
3.53 NA 8.29 0.58 1.46 12.51 1.98 1.35 29.70
1.36 NA 3.30 0.35 0.79 6.22 0.79 0.51 13.31
1.77 0.97 3.97 6.71 35.26
0.87 0.41 1.46 2.74 13.41
0.46 1.68 3.15 5.29 12.03
0.68 2.14 3.49 6.31 14.35
0.28 0.94 5.06 6.27 17.84
0.19 0.54 2.54 3.27 9.33
1.87 2.20 9.13 13.20 42.90
1.04 1.05 3.80 5.89 19.21
Note: The IT infrastructure indicator for Estonia is higher than the average of the EU-15. We do not display trade gains here.
Table 2. Relative trade gains of new and candidate EU members– Raising capacity half-way to the EU Level Country New members Czech Republic Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia Sub total average: Candidate members Bulgaria Romania Turkey Sub total average: AVERAGE:
Port efficiency
Customs regimes
Regulatory policy
IT infrastructure
13.40% 3.34% 17.46% 5.03% 10.31% 12.01% 29.56% 8.05% 12.40%
6.87% 1.30% 1.09% 10.22% 4.33% 9.01% 3.73% 2.94% 4.94%
5.54% 1.80% 7.25% 3.29% 2.28% 8.34% 6.06% 4.89% 4.93%
7.46% NA 18.73% 18.28% 26.13% 23.63% 11.37% 10.10% 16.53%
22.81% 5.71% 8.34% 12.29% 12.37%
9.89% 15.74% 10.19% 11.94% 6.85%
3.99% 6.11% 11.68% 7.26% 5.57%
25.16% 13.40% 19.86% 19.48% 17.41%
Note: The IT infrastructure development level of Estonia is higher than the average of the EU-15, so its trade gains are not displayed here.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
42 - TRADE AND TRANSPORT FACILITATION:
EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES
4. EXTERNALITIES FROM INVESTMENT IN TRADE FACILITATION
Trade facilitation improvements benefit not only the countries that implement them, but also their trading partners. The more intense are trade relations between countries, the greater the benefit partner countries will enjoy should one (or more) of the countries improve its trade facilitation regime, all others being equal. Given the importance of intra-regional trade between the EU-15 and the new and candidate EU members, the expected gains to the former resulting from the trade facilitation improvements in the latter will likely be significant. Table 3 shows that if all four dimensions of trade facilitation for each of the new member (EU-8) and candidate member countries (Bulgaria, Romania and Turkey) improve up to half the EU-15 level, almost $10 billion in trade gains will be expected for the EU-15, the EU-8 and the three candidate members as a whole in 20018. Among the total gains for trading partner countries, $2.7 billion (or 28 per cent) comes from their export gains and $6.9 billion (or 72 per cent) from their import gains (Table 3, Table 4). In other words, if the trade facilitation capacity of new and candidate EU members improves measurably, their trade volumes with other European countries will also increase; in particular, partner countries will increase their volumes of imports from the improving countries by a larger margin than partner exports increase to those countries9. These findings echo our argument that trade facilitation improvements in new and candidate EU member countries will strongly increase their intra-regional trade with the EU-15. The trade gains for the EU-15 represent 74 per cent of the total trade gains for all European countries as a whole, among which the import gains for the EU-15 represent 54 per cent of the total gains. Among the four dimensions of trade facilitation, improvement of IT infrastructure will result in the highest trade gains (more than $4 billion), which is greater than the gains from improvement in port efficiency (with trade gains approaching $3 billion). Improvements in regulatory policy and customs regimes share quasi-equally trade gains of around $1.5 billion each. In other words, more than 40 per cent of the trade gains come from improvements in IT infrastructure and almost 30 per cent from port efficiency. This suggests that a higher priority for improvement be accorded to IT infrastructure and port efficiency for new and candidate EU members.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES -
43
Table 3. Absolute trade gains of European partner countries resulting from trade facilitation improvements of the new and candidate member countries (unit: million $) Candidate member countries Export gain Import gain Export gain Import gain Export gain Import gain EU-15
Port efficiency: New members Candidate member countries EU8+3 candidate members Customs regimes: New members Candidate member countries EU8+3 candidate members Regulatory policy: New members Candidate member countries EU8+3 candidate members IT infrastructure: New members Candidate member countries EU8+3 candidate members TOTAL:
New members
370.64 73.63 444.27
1089.14 398.38 1487.52
153.44 16.19 169.63
476.88 38.54 515.42
17.01 8.64 25.65
93.39 33.31 126.70
337.10 180.97 518.07
312.21 292.62 604.84
110.78 55.99 166.77
105.81 28.41 134.22
12.57 21.75 34.32
14.85 19.56 34.41
221.96 87.91 309.87
481.56 384.13 865.69
71.20 18.95 90.15
166.32 27.38 193.70
10.14 8.40 18.54
35.43 19.00 54.43
519.95 159.42 679.37 1951.57
1430.81 789.33 2220.14 5178.18
167.69 36.45 204.14 630.69
449.14 66.82 515.96 1359.29
23.40 17.22 40.62 119.14
120.50 51.79 172.29 387.84
Table 4. Relative trade gains of European partner countries resulted from trade facilitation improvements of the new member and candidate member countries Candidate member countries Export gain Import gain Export gain Import gain Export gain Import gain EU-15
Port efficiency: New members Candidate member countries EU8+3 candidate members Customs regimes: New members Candidate member countries EU8+3 candidate members Regulatory policy: New members Candidate member countries EU8+3 candidate members IT infrastructure: New members Candidate member countries EU8+3 candidate members TOTAL:
New members
3.85% 0.76% 4.61%
11.31% 4.14% 15.45%
1.59% 0.17% 1.76%
4.95% 0.40% 5.35%
0.18% 0.09% 0.27%
0.97% 0.35% 1.32%
3.50% 1.88% 5.38%
3.24% 3.04% 6.28%
1.15% 0.58% 1.73%
1.10% 0.30% 1.39%
0.13% 0.23% 0.36%
0.15% 0.20% 0.36%
2.31% 0.91% 3.22%
5.00% 3.99% 8.99%
0.74% 0.20% 0.94%
1.73% 0.28% 2.01%
0.11% 0.09% 0.19%
0.37% 0.20% 0.57%
5.40% 1.66% 7.06% 20.27%
14.86% 8.20% 23.06% 53.79%
1.74% 0.38% 2.12% 6.55%
4.67% 0.69% 5.36% 14.12%
0.24% 0.18% 0.42% 1.24%
1.25% 0.54% 1.79% 4.03%
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
44 - TRADE AND TRANSPORT FACILITATION:
EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES
Tables 5a-d detail the trade gains for the EU-15, EU-8 and three candidate member countries, resulting from different types of trade facilitation improvements by each new and candidate member country in absolute terms. It is not surprising that there are large trade gains from trade facilitation improvements by the Czech Republic, Hungary and Poland, given their large economic size. The larger the improvement one country makes in trade facilitation, the larger the trade gains. For a large country like Poland, which the data suggest requires significant improvements in trade facilitation, considerable progress is required to approach the half-way level of the EU-15.
Table 5a. Absolute trade gains of European partner countries resulting from port efficiency improvement (half-way to EU-15) of each new member and candidate member country (unit: million $) Port efficiency New members: Czech Republic Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia Subtotal: Candidate members: Bulgaria Romania Turkey Subtotal: TOTAL:
Candidate member countries Export gain Import gain Export gain Import gain Export gain Import gain EU-15
New members
1 445.73 46.59 1 644.67 51.10 166.00 1 880.95 1 068.83 249.54 6 553.43
4 254.22 122.28 5 442.06 104.90 300.69 4 141.78 3 096.13 625.13 18 087.20
228.14 7.32 165.69 20.12 53.45 212.69 419.32 28.39 1 135.11
928.23 22.14 536.00 34.45 69.88 552.59 1 235.44 87.41 3 466.12
16.64 0.49 46.82 0.76 1.89 34.39 14.21 4.43 119.63
102.77 0.28 216.90 0.08 14.05 95.11 99.19 17.19 545.56
345.14 260.93 735.15 1 341.22 7 894.65
977.98 719.21 2 198.22 3 895.41 21 982.62
39.71 39.33 27.75 106.80 1 241.91
56.93 47.92 119.22 224.06 3 690.19
30.13 13.37 21.53 65.03 184.65
94.94 35.79 69.61 200.34 745.90
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES -
45
Table 5b. Absolute trade gains of European partner countries resulting from customs regimes improvement (half-way to EU-15) of each new member and candidate member country (unit: million $) Customs regimes New members: Czech Republic Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia Subtotal: Candidate members: Bulgaria Romania Turkey Subtotal: TOTAL:
Candidate member countries Export gain Import gain Export gain Import gain Export gain Import gain EU-15
New members
1 477.70 34.19 203.11 184.39 127.25 2 631.62 266.40 182.31 5 106.97
1 444.72 29.81 223.30 125.76 76.58 1 925.30 256.40 151.74 4 233.62
233.18 5.37 20.46 72.58 40.97 297.57 104.52 20.74 795.39
315.22 5.40 21.99 41.30 17.80 256.87 102.31 21.22 782.11
17.01 0.36 5.78 2.74 1.45 48.11 3.54 3.23 82.22
34.90 0.07 8.90 0.10 3.58 44.21 8.21 4.17 104.14
270.77 1 356.64 1 755.39 3 382.79 8 489.76
254.92 1 242.38 1 743.95 3 241.25 7 474.87
31.16 204.50 66.27 301.93 1 097.32
14.84 82.77 94.58 192.19 974.30
23.64 69.49 51.40 144.53 226.76
24.75 61.83 55.22 141.80 245.93
Table 5c. Absolute trade gains of European partner countries resulting from regulatory policy improvement (half-way to EU-15) of each new member and candidate member country (unit: million $) Regulatory policy New members: Czech Republic Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia Subtotal: Candidate members: Bulgaria Romania Turkey Subtotal: TOTAL:
Candidate member countries Export gain Import gain Export gain Import gain Export gain Import gain EU-15
New members
746.96 30.99 850.51 40.54 44.84 1 602.45 272.83 189.25 3 778.36
1 611.31 59.62 2 063.06 61.00 59.55 2 586.69 579.37 347.55 7 368.15
117.87 4.87 85.68 15.96 14.44 181.20 107.04 21.53 548.58
351.57 10.79 203.19 20.03 13.84 345.11 231.18 48.60 1 224.32
8.60 0.33 24.21 0.60 0.51 29.29 3.63 3.36 70.53
38.92 0.14 82.23 0.05 2.78 59.40 18.56 9.56 211.63
73.46 343.71 1 279.26 1 696.43 5 474.79
152.59 694.49 2 804.18 3 651.26 11 019.41
8.45 51.81 48.30 108.56 657.14
8.88 46.27 152.08 207.23 1 431.55
6.41 17.60 37.46 61.48 132.01
14.81 34.56 88.80 138.17 349.80
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
46 - TRADE AND TRANSPORT FACILITATION:
EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES
Table 5d. Absolute trade gains of European partner countries resulted from IT infrastructure improvement (half-way to EU-15) of each new member and candidate member country (unit: million $) IT infrastructure New members: Czech Republic Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia Subtotal: Candidate members: Bulgaria Romania Turkey Subtotal: TOTAL:
Candidate member countries Export gain Import gain Export gain Import gain Export gain Import gain EU-15
New members
880.45 NA 1 926.69 200.72 456.31 4 020.49 449.16 342.45 8 276.27
2 294.29 NA 5 645.56 364.88 731.97 7 839.71 1 152.18 759.69 18 788.27
138.94 NA 194.10 79.01 146.92 454.62 176.21 38.96 1 228.77
500.59 NA 556.04 119.82 170.10 1 045.96 459.75 106.23 2 958.49
10.13 NA 54.85 2.98 5.19 73.50 5.97 6.07 158.71
55.42 NA 225.02 0.28 34.19 180.02 36.91 20.89 552.73
412.26 666.29 1 911.45 2 990.01 11 266.28
1 034.49 1 626.30 5 061.39 7 722.19 26 510.45
47.44 100.44 72.16 220.04 1 448.81
60.22 108.35 274.50 443.07 3 401.55
36.00 34.13 55.97 126.10 284.80
100.43 80.93 160.27 341.63 894.36
Tables 6a-d show the corresponding relative trade gains. The results show that, in contrast to the large absolute trade gains that the EU-15 will enjoy, due to improvements in trade facilitation by the EU-8 and candidate member countries it is the two latter groups of countries that will benefit the most relative to their own trade volumes. In particular, the relative trade gains of the partner EU-8 and candidate member countries are quite large, should the three largest of these economies (the Czech Republic, Hungary and Poland) improve their trade facilitation. For example, if Poland increases its IT infrastructure to half the EU-15 average, the other seven new member countries will enjoy a trade gain of 0.8 per cent (of which the export gain is 0.5 per cent and the import gain is 1.0 per cent), and the three candidate member countries will enjoy a trade gain of 0.25 per cent (of which the export gain is 0.16 per cent and the import gain is 0.33 per cent), while the EU-15 enjoy a trade gain of 0.29 per cent (of which the export gain is 0.19 per cent and the import gain is 0.39 per cent). This suggests that trade facilitation improvements by the new and candidate EU members will largely increase the trade volumes among themselves disproportionately, thanks to the relatively intense trade relationships among them.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES -
47
Table 6a. Trade gains of European partner countries (relative to their respective total trade volume) resulting from port efficiency improvement (half-way to EU-15) of each new member and candidate member country Port Efficiency
Czech Rep. Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia Bulgaria Romania Turkey
EU-15 Export gain 6.75 0.22 7.68 0.24 0.78 8.79 4.99 1.17 1.61 1.22 3.43
Import gain 21.30 0.61 27.25 0.53 1.51 20.74 15.50 3.13 4.90 3.60 11.01
New members Trade gains 13.77 0.41 17.13 0.38 1.13 14.55 10.06 2.11 3.20 2.37 7.09
Export gain 24.16 0.60 17.12 1.61 4.34 22.94 36.47 2.41 3.13 3.10 2.19
Candidate member countries
Import gain 82.94 1.58 47.33 2.43 4.99 55.31 93.18 6.44 3.92 3.30 8.22
Trade gains 56.04 1.12 33.41 2.05 4.68 39.73 66.84 4.57 3.55 3.21 5.41
Export gain 3.60 0.11 10.14 0.16 0.41 7.44 3.08 0.96 7.25 3.76 14.04
Import gain 18.78 0.05 39.64 0.01 2.57 17.38 18.13 3.14 19.86 8.70 33.98
Trade gains 11.83 0.08 26.13 0.08 1.58 12.83 11.24 2.14 14.00 6.41 25.44
Note: Figures are presented in 1/100000. For example, if the port efficiency development level of the Czech Republic reaches half of the EU-15 average, the export volume of EU-15 will increase 0.000675.
Table 6b. Relative trade gains of European partner countries (relative to their respective total trade volume) resulting from customs regimes improvement (half-way to EU-15) of each new member and candidate member country Customs regimes
Czech Rep. Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia Bulgaria Romania Turkey
EU-15 Export gain 6.90 0.16 0.95 0.86 0.59 12.29 1.24 0.85 1.26 6.34 8.20
Import gain 7.23 0.15 1.12 0.63 0.38 9.64 1.28 0.76 1.28 6.22 8.73
New members Trade gains 7.06 0.15 1.03 0.75 0.49 11.01 1.26 0.81 1.27 6.28 8.46
Export gain 24.70 0.44 2.11 5.81 3.32 32.10 9.09 1.76 2.46 16.13 5.23
Import gain 28.17 0.38 1.94 2.91 1.27 25.71 7.72 1.56 1.02 5.70 6.52
Trade gains 26.58 0.41 2.02 4.27 2.23 28.78 8.35 1.66 1.69 10.57 5.92
Candidate member countries Export gain 3.68 0.08 1.25 0.59 0.31 10.41 0.77 0.70 5.69 19.56 33.52
Import gain 6.38 0.01 1.63 0.02 0.65 8.08 1.50 0.76 5.18 15.03 26.95
Trade gains 5.14 0.04 1.45 0.28 0.50 9.15 1.16 0.73 5.42 17.13 29.76
Note: Figures are presented in 1/100000. For example, if the customs regimes development level of the Czech Republic reaches half of the EU-15 average, the export volume of EU-15 will increase 0.0000690.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
48 - TRADE AND TRANSPORT FACILITATION:
EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES
Table 6c. Relative trade gains of European partner countries (relative to their respective total trade volume) resulting from regulatory policy improvement (half-way to EU-15) of each new member and candidate member country on their exports and imports Regulatory policy
Czech Rep. Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia Bulgaria Romania Turkey
EU-15 Export gain 3.49 0.14 3.97 0.19 0.21 7.49 1.27 0.88 0.34 1.61 5.98
Import gain 8.07 0.30 10.33 0.31 0.30 12.95 2.90 1.74 0.76 3.48 14.04
New members Trade gains 5.70 0.22 7.04 0.25 0.25 10.12 2.06 1.30 0.55 2.51 9.87
Export gain 12.48 0.40 8.85 1.28 1.17 19.55 9.31 1.83 0.67 4.09 3.81
Import gain 31.41 0.77 17.94 1.41 0.99 34.54 17.44 3.58 0.61 3.19 10.48
Candidate member countries Trade gains 22.75 0.60 13.75 1.35 1.07 27.32 13.66 2.77 0.64 3.61 7.37
Export gain 1.86 0.07 5.24 0.13 0.11 6.34 0.79 0.73 1.54 4.96 24.43
Import gain 7.11 0.02 15.03 0.01 0.51 10.85 3.39 1.75 3.10 8.40 43.34
Trade gains 4.71 0.05 10.55 0.06 0.33 8.79 2.20 1.28 2.38 6.80 35.24
Note: Figures are presented in 1/100000. For example, if the regulatory policy development level of the Czech Republic reaches half of the EU-15 average, the export volume of EU-15 will increase 0.0000349.
Table 6d. Relative trade gains of European partner countries (relative to their respective total trade volume) resulting from IT infrastructure improvement (half-way to EU-15) of each new member and candidate member country IT infrastructure
Czech Rep. Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia Bulgaria Romania Turkey
Export gain 4.11 NA 9.00 0.94 2.13 18.78 2.10 1.60 1.93 3.11 8.93
EU-15 Import gain 11.49 NA 28.27 1.83 3.66 39.25 5.77 3.80 5.18 8.14 25.34
Trade gains 7.67 NA 18.30 1.37 2.87 28.66 3.87 2.66 3.50 5.54 16.85
New members Export Import Trade gain gain gains 14.71 44.73 30.99 NA NA NA 20.06 49.10 35.72 6.33 8.44 7.45 11.92 12.15 12.04 49.04 104.69 77.90 15.33 34.68 25.69 3.31 7.82 5.73 3.74 4.15 3.96 7.92 7.47 7.68 5.69 18.92 12.75
Candidate member countries Export Import Trade gain gain gains 2.19 10.13 6.50 NA NA NA 11.87 41.12 27.73 0.65 0.05 0.32 1.12 6.25 3.90 15.91 32.90 25.12 1.29 6.75 4.25 1.31 3.82 2.67 8.67 21.00 15.27 9.61 19.67 15.01 36.50 78.23 60.37
Note: Figures are presented in 1/100000. For example, if the IT infrastructure development level of the Czech Republic reaches half of the EU-15 average, the export volume of EU-15 will increase 0.000411.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES -
49
5. CONCLUDING REMARKS – INFORMING CAPACITY-BUILDING POLICY PRIORITIES
Our results suggest that trade gains are positively correlated with the development of behind-theborder trade facilitation capacity. Of course, projected trade gains depend on a number of other factors, such as sensitivity of trade volumes to changes in specific behind-the-border barriers, and the overall structure of trade of the country in question, for instance. Moreover, the results presented here are subject to data constraints and this analysis provides a first step in understanding the complexities of trade facilitation in a broader context. New data sets are key to future research work in this area. Building on analysis in Wilson et al. (2004), the analysis here suggests that improvement in IT infrastructure has the greatest marginal impact on trade gains among the four dimensions of trade facilitation. A long, paper-based system in trade transactions is highly inefficient. The lack of co-operation between multiple-border agencies results in information inconsistencies. Excessive bureaucracy, lack of transparency and the ambiguity in regulatory interpretation can also lead to unnecessary transaction costs. If clearance procedures could be streamlined, the attendant time would be shortened and costs would be saved. According to estimates prepared under the World Bank "Trade and Transport Facilitation in Southeast Europe” (TTFSE) project, the current long waiting times at border crossings are highly inefficient and costly. If the current crossing time in Bulgaria is reduced by 25 minutes, for example, the savings on transport costs are estimated in excess of US$5 million per year after 2005 for vehicles entering Bulgaria. What improvements in clearance capabilities are likely to have significant payoffs? The adoption of international Electronic Data Interchange standards will enhance border efficiency by processing the e-information before actual docking and enable fast integration into the international trading community. It not only cuts down the waiting time but also enables centralised audits for valuation of traded goods by reducing information inconsistencies. The creation of databases that are accessible in real time at all customs locations will support selectivity and targeting procedures, and will reduce corruption opportunities by lessening human intervention and by monitoring customs performance. The streamlining of border controls, such as the implementation of a single payment window at border crossing points and the paperless transactions for customs clearance could result in large cost savings. The significant costs associated with the observed deficits in trade facilitation capacities in the new and candidate member countries of the EU, engender strong incentives for them to reallocate resources so as to achieve the maximum impact from improvements in such capacities on trade gains and economic growth. The analysis in this paper suggests that priorities for capacity building differ with the two sets of countries examined. Among the four trade facilitation indicators, IT infrastructure improvement will lead to the largest gain for the new member countries, as well as for candidate member countries. Figure 5 shows that, if all indicators improve to half of the EU-15 level, almost 40 per cent trade gains will result from the improvement of IT infrastructure. As to the trade gains that could result from improvement in the other trade facilitation activities examined, however, they differ between the new and candidate member countries. The new members TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
50 - TRADE AND TRANSPORT FACILITATION:
EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES
of the European Union exhibit large potential gains to trade with investment in port efficiency (both air and maritime ports), which represents a third of the total trade gain. The results for the candidate members suggest more widely dispersed gains, with investments in port efficiency, customs regimes and regulatory policy. Improvements in each dimension share around 20 per cent of the total trade gains.
Figure 5. Relative trade gains due to the improvement of each trade facilitation indicator
5a – EU-8
5b - 3 candidate members
Development planning and capacity-building programmes can take these results into consideration – along with many other factors – as integration into the European Union continues. In general, improvements in port facilities and IT infrastructures are likely to be more costly than the administrative reforms at the centre of customs regimes and regulatory policy – but they can have correspondingly high payoffs. The eligibility for additional EU financing with accession should provide more scope for improvements in these areas. For the candidate member countries that are currently negotiating the formal accession with the EU, such as Bulgaria and Romania, it may be of particular interest to explore acceleration of telecommunications liberalisation and investment, for example.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES -
51
ANNEX 1 Summary Statistics for Values of Trade Facilitation Indicators Category Port Efficiency
Indexed inputs
Source
Mean
Std. Dev.
Min.
Ports Facilities
GCR
.636
.189
.261
Air Transport
GCR
.710
.166
.673
GCR GCR
Aggregate Index Customs Hidden import barriers Regimes Bribery Aggregate Index Transparency of Regulatory government policies Policy Control of corruption Aggregate Index Information Speed and costs of Technology Internet access Infrastructure Effect of Internet on business Aggregate Index
Min. Importer Max.
Max. Importer
1.000
Singapore
.229
Bolivia Slovak Republic
1.000
Singapore
.169
.345
Bolivia
1.000
Singapore
.702
.167
.368
Paraguay
1.000
Finland
.689
.175
.343
Bangladesh
1.000
Iceland
.695
.163
.384
Paraguay
0.979
Finland
WCY
.619
.205
.089
Argentina
1.000
Finland
KKZ
.746
.140
.530
South Africa
1.000
Finland
.689
.139
.353
Venezuela
1.000
Finland
GCR
.629
.162
.348
Vietnam
1.000
Finland
GCR
.719
.102
.481
Greece
1.000
Finland
.674
.121
.482
Mauritius
1.000
Finland
Source: Wilson, Mann and Otsuki (2004).
Data are from the World Economic Forum, Global Competitiveness Report, 2001-02 (GCR), IMD Lausanne, World Competitiveness Yearbook 2000 (WCY), and Kaufmann, Kraay and Zoido-Lobaton (2002) (KKZ). All survey data in GCR comes from the World Economic Forum’s Executive Opinion Survey. A total of 4 022 firms were surveyed. “In order to provide the basis for a comparative assessment on a global basis, it is essential that we interview a sufficient number of senior business leaders in individual countries and that the sample in each country is not biased in favor of any particular business group. We have taken a number of steps to ensure this. First, we have asked each of our partner institutes, the organizations that administer the surveys in each country, to start with a comprehensive register of firms. From this, they were asked to choose a sample whose distribution across economic sectors was proportional to the distribution of the country’s labor force across sectors, excluding agriculture. They were then asked to choose firms randomly within these broad sectors (for example, by choosing firms at regular intervals from an alphabetic list), and to pursue face-to-face interviews, following up for clarifications where necessary. The employment distribution was taken from data in the 1998 Yearbook of Labour Statistics of the International Labour Office. The respondents to the survey are typically a company’s CEO or a member of its senior management.”
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
52 - TRADE AND TRANSPORT FACILITATION:
EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES
The WCY uses a 115-question survey, sent to executives in top and middle management of firms in all 49 countries of the WCY. The sample size of each country is proportional to GDP, and firms "normally have an international dimension." The firms are selected to represent a cross-section of manufacturing, service and primary industries. There were 3 532 responses to the survey. KKZ (2002) updates the data on governance that were developed in Kaufmann, Kraay and Zoido-Lobaton (1999), “Governance Matters.” The database contains more than 300 governance indicators for 175 countries, compiled from a variety of sources in 2000-2001. Six aggregate indicators are constructed, corresponding to six basic governance concepts: Voice and Accountability, Political Stability, Government Effectiveness, Regulatory Quality, Rule of Law and Control of Corruption. The various raw data series were chosen because of their relevance to the four concepts of trade facilitation:
Port efficiency” for each country J is the average of two indexed inputs (all GCR): o o
“Customs Regimes” for each country J is the average of two indexed inputs (all GCR): o o
Hidden import barriers other than published tariffs and quotas; Irregular extra payments or bribes connected with import and export permits.
“Regulatory policy” for each country J is constructed as the average of four indexed inputs: o o
Port facilities and inland waterways are: (1 = underdeveloped, 7 = as developed as the world's best, GCR); Air transport is: (1 = infrequent and inefficient, 7 = as extensive and efficient as the world's best, GCR).
Transparency of government policy is satisfactory (WCY); Control of Corruption (KKZ).
“Information technology infrastructures” for each country J is as the average of three indexed inputs (all GCR): o o
Speed and cost of Internet access are: (1=slow and expensive, 7 = fast and cheap); Internet contribution to reduce inventory costs is: (1 = no improvement, 7 = huge improvement).
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES -
53
ANNEX 2 Description of the Gravity Model in Wilson, Mann and Otsuki (2004)10
Wilson, Mann and Otsuki (2004) develop the gravity equation to examine the role of trade facilitation, measured by port efficiency, customs regimes, regulatory policy and information technology infrastructure, in the determination of bilateral trade: ln(VJIt ) = b1ln(100+TARIFFJIt) + b2 lnPEJ + b3lnREJ + b4lnSIJ + b5lnPEI + b6lnCEI + b7lnREII + b8lnSII + b9ln(GNPIt) +b10ln(GNPJt) + b11ln(GNPPCIt)+ b12ln(GNPPCJt) + b13ln(DISTIJ)+b14DADJ +b15 DASEAN + b16DNAFTA + b17 DLAIA + b18 DAUNZ + b19 DMERCOSUR + b20DEU + b21DENG + b22DFRC + b23DSPN + b24 DARB + b25 DCHN + b26 DGMN + b27 DPOR + b28 DRUS + b29 D2000 + ε JI
t
where I and J stand for the importer and exporter respectively, and t denotes trading years (t=2000, 2001). Parameters b are coefficients. The term ε JI t is the error term, assumed to be normally distributed with mean zero. The value of manufactures exported from country J to I is denoted as VJI (so exporter to importer). The term TARIFFJI denotes applied tariff rate in the per cent ad valorem term that is specific to the trading partners I and J and year t. The inclusion of the tariff variable is useful for reducing omitted variable biases. It is particularly important for some nations since, unlike the EU whose tariff policies are harmonized, applied tariff rates generally vary across most other countries and possibly across their exporting partners. The terms PEJ, REJ and SIJ denote exporting country J’s indicators of port efficiency, regulatory policy and information technology infrastructures. Similarly, PEI, REI and SII stand for the same trade facilitation measures in the importing country. For the importing country, we include one additional measure, i.e. “Customs Regimes”, or CEI. We use “Customs Regimes” only for the importers, since in bilateral trade customs is more relevant as a factor affecting imports than exports. This set of trade facilitation variables is different than in WMO. There, we included only PEI, REI, SII , and CEI. That is, for country I we considered only the effect on imports of unilateral trade facilitation. Country I’s exports improved indirectly when its trading partners improved their trade facilitation efforts. In this formulation, we take explicit account of the fact that country J’s exports (as well as its imports) will improve through its own trade facilitation efforts. The term GNP denotes Gross National Product and GNPPC denotes per capita GNP, where both are expressed in 1995 US dollar terms. The geographical distance between capital cities I and J is denoted as DISTIJ. Dummy variables are included to capture the effect of preferential trade arrangements, language similarity and adjacency. The trade arrangements dummies include NAFTA (DNAFTA), ASEAN (DASEAN), LAIA (DLAIA), AUNZ (DAUNZ), MERCOSUR (DMERCOSUR) and EU (DEU). TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
54 - TRADE AND TRANSPORT FACILITATION:
EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES
The language dummies include English (DENG), French (DFRC), Spanish (DSPN), Arabic (DARB), Chinese (DCHN), German (DGMN), Portuguese (DPOR) and Russian (DRUS). The adjacency dummy DADJ takes the value of one if country I is adjacent to country J and zero otherwise. Additionally, a dummy for year 2000 is included in the model to control for time-specific shocks. The estimation results are as follows: Coefficent Constant Tariff rates Port efficiency of importer Port efficiency of exporter Customs environment of importer Regulatory environment of importer Regulatory environment of exporter Service-sector infrastructures of importer Service-sector infrastructures of exporter GNP of importer Per capita GNP of importer GNP of exporter Per capita GNP of exporter Geographical distance Adjacency dummy Year 2000 dummy Adjusted R-squared Number of observations
-10.641*** -1.155*** 0.307* 0.924*** 0.472** 0.281* 0.620*** 0.729*** 1.943*** 0.915*** -0.182*** 1.246*** -0.226*** -1.258*** 0.336*** -0.031 0.758 7 904
Standard Error 1.558 0.318 0.163 0.148 0.199 0.144 0.132 0.224 0.216 0.014 0.037 0.014 0.029 0.025 0.114 0.039
Note: The significance levels at 10%, 5% and 1% are denoted by “*”, “**”, and “***”, respectively. For the sake of simplicity, the estimations of the geographic dummies are not shown in the table.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRADE AND TRANSPORT FACILITATION: EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES -
55
NOTES
1.
The data on port efficiency, customs regimes, regulatory policy and information technology infrastructures for Cyprus, Malta and Croatia are not available. Given the relatively small economic size of Cyprus and Malta, we focus on the study of the other eight new member countries of EU. Data for Croatia are not currently available. .
2.
See Annex 1 for a summary of the statistics and data sources of the four trade facilitation indicators.
3.
Due to the data constraint, we could not measure the efficiency of the port facilities and inland waterways separately.
4.
The data for Luxembourg is not available. We use the average of the other 14 member countries as being that of the EU-15.
5.
Hungary and Estonia share the similar development level of customs regimes.
6.
For the sake of comparison, the real GDP growth rate of Hungary was 2.8% in 2003 (data source: CIA Fact Book).
7.
Many of the new member and candidate countries have a two-digit unemployment rate. For example, Poland and the Slovak Republic have unemployment rates over 15 per cent, the Czech Republic, Estonia and Latvia over 10 per cent.
8.
Here, we use the data on the bilateral import volume of Bulgaria from the EU-15, other EU-8 countries, Romania and Turkey in 2000, since the 2001 data is not available. As argued above, although the IT infrastructure development level of Estonia is better than the EU-15 average, we neglect the IT infrastructure improvement in Estonia. That is to say, we suppose that no improvements of IT infrastructure occur in Estonia. When we simulate the trade gain of an EU-8 or a candidate member country resulting from other EU-8 or candidate member countries, we neglect the trade gain of the country in question. In other words, we simulate the trade gain on the other EU-8 countries (or other candidate member countries) resulting from the improvement of trade facilitation in the new or candidate member country in question.
9.
The reason for the higher import gains for partner countries stems from a higher elasticity of export to trade facilitation improvement than the corresponding elasticity for imports, as found by WMO (2004). This implies that if a country improves its trade facilitation, its export volume increase will be higher than its import volume increase.
10. Source: Wilson, Mann and Otsuki (2004).
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
56 - TRADE AND TRANSPORT FACILITATION:
EUROPEAN ACCESSION AND CAPACITY BUILDING PRIORITIES
BIBLIOGRAPHY
Overman, H.G., S. Redding and A.J. Venables (2001), “The Economic Geography of Trade, Production, and Income: A Survey of Empirics”, http://econ.lse.ac.uk/staff/ajv/hosrtv.pdf Diamond, J. (1997), Guns, Germs, and Steel: The fates of Human Societies, New York: W.W. Norton. Limao, N. and A.J. Venables (2001), “Infrastructure, Geographical Disadvantages, Transport Costs and Trade”, World Bank Economic Review, 15, pp. 451-479. Redding, S. and A.J. Venables (2003), “Geography and Export Performance: External Market Access and Internal Supply Capacity”, website: http://econ.lse.ac.uk/staff/ajv/Isit22.pdf Hummels, D. (1999), “Have International Transportation Costs Declined?”, mimeo, http://www.mgmt.purdue.edu/faculty/hummelsd/ IMD (2000), World Competitiveness Yearbook, IMD, Lausanne. Kaufmann, D., A. Kraay and P. Zoido-Lobaton (2002), Governance Matters II: Updated Indicators for 2000-01, World Bank Working Paper No 2772, The World Bank, Washington, DC. McCallum, J. (1995), "National border matters: Canada-US regional trade patterns", American Economic Review, Vol. 85, No. 3, pp. 615-623. Wilson, J.S., C.L. Mann and T. Otsuki (2004), Assessing the Potential Benefit of Trade Facilitation: A Global Perspective, World Bank Working Paper Series, No. 3224, The World Bank, Washington, DC. Wilson J.S., C.L. Mann, Y. Woo, N. Assanie and I. Choi (2002), Trade Facilitation: A Development Perspective in the Asia-Pacific Region, Asia Pacific Economic Cooperation, Singapore. World Bank (2004), "Framework for World Bank Support to the European Union (EU) New Member Countries of Central and Eastern Europe", World Bank mimeo 29135. World Economic Forum (2000), Global Competitiveness Report, World Economic Forum, Geneva. Woo, Yuen Pau and John S. Wilson (2000), Cutting Through Red Tape: New Directions for APEC's Trade Facilitation Agenda, Asia Pacific Foundation of Canada: Vancouver.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
Anthony VENABLES London School of Economics and Political Science London United Kingdom
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
- 57
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
- 59
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
SUMMARY
1.
INTRODUCTION ......................................................................................................................... 61
2.
TRADE COSTS AND REAL INCOME: ANALYSIS................................................................. 61 2.1. Perfect competition and inter-industry trade.......................................................................... 62 2.2. Increasing returns to scale and intra-industry trade................................................................ 66 2.3. Other mechanisms.................................................................................................................. 70
3.
TRADE AND REAL INCOME: EMPIRICS................................................................................ 71 3.1. 3.2. 3.3. 3.4.
4.
Measuring market access ....................................................................................................... 71 Supplier access and prices...................................................................................................... 72 Market access and real income............................................................................................... 72 Market access and income: conclusions................................................................................. 77
TRADE COSTS, TRADE FLOWS AND INFRASTRUCTURE................................................. 78 4.1. Determinants of trade costs.................................................................................................... 78 4.2. Determinants of trade flows ................................................................................................... 80 4.3. The role of infrastructure ....................................................................................................... 80
5.
CONCLUDING COMMENTS ..................................................................................................... 81
NOTES .................................................................................................................................................. 82 ANNEX .................................................................................................................................................. 83 BIBLIOGRAPHY ................................................................................................................................. 84
London, August 2004
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
- 61
1. INTRODUCTION
How does transport infrastructure affect countries’ ability to participate in the international economy, and thereby affect their real income and development prospects? This paper explores some economic approaches to answering these questions, drawing on both theoretical and empirical work on international trade and economic geography. It does not seek to provide a comprehensive overview of literature, but instead to outline some of the effects that operate, and discuss some attempts that have been made to quantify these effects. Answering the questions posed above involves two steps. One is: how does a change in transport infrastructure affect the costs of international trade? The other is: how does a change in the cost of trade affect a country, and what impact does it have on a country’s trade and real income? This paper starts off – and is mainly devoted to – the second of these questions. While the gains from trade are widely accepted, less is known about the magnitude of the penalty faced by countries for which trade is costly. Reducing trade costs has direct benefits and also leads to widespread adjustment in the economy: specialisation increases and factor prices and real incomes will change. How do these effects operate, and how large might they be? Chapter 2 of the paper presents a theoretical review of some possibilities, drawing on different approaches in the international economics and economic geography literatures. Chapter 3 turns from theory to empirical studies, reporting attempts that have been made to quantify the real income penalty associated with poor market access. There is evidence of substantial “wage gradients”, with real income steadily declining from “central” countries, with good market access, to “peripheral” ones. Chapter 4 investigates the relationships between infrastructure, trade costs and trade volumes. A particular infrastructure investment will affect trade costs on a transport link in a detailed manner that is beyond the scope of this paper. However, there are studies based on highly aggregate indices of infrastructure quality that attempt to quantity the importance of infrastructure, and demonstrate that it has a major impact on trade costs and trade volumes. We report some of these estimates, and thereby provide some indication of the extent to which improvements in infrastructure might be associated with lower trade costs and increased trade volumes. We link these estimates to the empirical evidence of Chapter 3 to produce some – highly speculative – estimates of the value of infrastructure improvements. Finally, Chapter 5 offers some concluding remarks about the implications of findings for methodologies used in transport appraisal.
2. TRADE COSTS AND REAL INCOME: ANALYSIS
International trade incurs a variety of real costs. These include the costs of time in transit and of uncertainty in delivery time, as well as direct freight costs. For analytical purposes we suppose that TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
62 - INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE these can all be aggregated into a simple monetary measure, like freight costs (but see the comments in subsection 2.1.2). How do such trade costs affect the real income of trading nations? They have a negative direct effect, similar to a worsening of the terms of trade, as countries have to use more real resources to purchase imports and as they receive less (net) for their exports. They also have indirect effects by changing the structure of production and factor prices in an economy, and causing countries to forego some of the gains from trade. This chapter outlines several related but distinct analytical approaches to thinking about these real income costs. The starting point is a standard trade framework in which all sectors are perfectly competitive and subject to non-increasing returns to scale (section 2.1). In this case transport costs restrict the extent to which countries can exploit comparative advantage, forcing them to produce domestically goods that could be produced more efficiently elsewhere. The models outlined are insightful for thinking about developing country issues, where countries do not have the technological capability to produce efficiently in many sectors. We then turn, in section 2.2, to the framework of “new trade theory” and economic geography. These models are designed to focus on intra-industry trade between countries with similar technological capabilities. Firms are footloose, and trade costs are important as they affect the attractiveness of countries as potential sites for firms’ investments. Countries with high trade costs will have relatively low wages to compensate firms for the “market access” disadvantage that they face.
2.1. Perfect competition and inter-industry trade 2.1.1
Trade costs and wage gradients
The point of departure is to consider a country which exports a good that it produces using domestic resources and imported inputs. The export industry is perfectly competitive, and a price taker on world markets. However, trade costs are incurred both on exporting the final good and on importing intermediates or capital equipment; so the domestic producer receives the world price minus shipping costs for output, and pays the world price plus shipping costs for imported inputs. How do these shipping costs affect the domestic value-added earned on the activity? The first point to make is that if value added is squeezed from both sides – lower output price and higher input costs – then quite small trade costs can have dramatic effects on domestic value added. An illustrative case is shown in Figure 1. Shipping costs are given on the horizontal axis, and are measured as a “trade cost factor”; this is essentially the c.i.f. price divided by the f.o.b. price. Thus, 1.0 corresponds to perfectly free trade and 1.5 is shipping costs equalling 50 per cent of the value of output. The vertical axis is domestic value added. Suppose that the good under consideration has 50 per cent of costs attributable to domestic value added (either in the sector directly, or value added in domestically produced intermediate goods used in the sector) and the other 50 per cent of costs attributable to imported inputs. (These shares are exact when production occurs at world prices but vary with prices; the production function is constant elasticity of substitution with elasticity 0.67; details are given in annex). What happens to domestic value added as trade costs increase? Trade costs of 10 per cent reduce domestic value added by 25 per cent; trade costs of 28 per cent (the median value of bilateral trade costs, according to the IMF cif/fob measure) reduce domestic value added by 60 per cent. What part of domestic value added bears this squeeze? It is not likely to be capital costs, probably higher in countries with poor infrastructure than in high-income countries. The lower dashed line gives the case where (at world prices) 50 per cent of domestic value added goes to capital and 50 per cent to labour, and where labour has to bear all the reduction in value added; trade costs of
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
- 63
28 per cent then reduce domestic wages to just 10 per cent of those they would be if the activity could operate at world prices. Figure 1. Trade costs and wages
Clearly, this is just an illustrative example, but it points to a harsh arithmetic. The final incidence of trade costs falls on domestic factors of production that are not internationally tradable – essentially labour. Labour may be only a small fraction of total costs but it has to bear all the trade costs, implying a large multiplier effect from trade costs to nominal wages. 2.1.2
Comparative advantage: factor intensity and transport intensity
The size of these effects depends on what we can term the “transport intensity” of the country’s export activities. The examples above were constructed for a case in which the export sector was quite “transport intensive”, bearing heavy shipping costs on both its output and its inputs. In fact, potential export activities – as diverse as electronic information services and cement production – differ widely in their “transport intensity”. Does this suggest that the examples above were overstated? Standard trade theory offers a theory of comparative advantage based on the interaction between the factor intensity of goods and the factor abundance of countries. In a world of trade costs, comparative advantage is also based on a further interaction between the “transport intensity” of goods and the transport costs of countries: see Venables and Limao (2002) for detailed analysis. Thus, a remote country with a suitably skilled labour force would produce transport un-intensive goods. If there is sufficient world demand for goods of this type then the analysis of the preceding subsection would be irrelevant; countries facing high trade costs could find suitable export goods which have low trade costs. In practise, however, it does not seem to be the case that “remote” countries have been successful in specialising in transport un-intensive goods. A lot of unskilled, labour-intensive products also seem to be quite transport intensive. For example, one of the fastest growing branches of world trade has TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
64 - INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE been trade in parts and components as firms outsource various stages of the production process. While individual components may be cheap to transport, this sort of activity is highly dependent both on imported inputs and on exporting its output; it is therefore “transport intensive” and, exactly as we saw in the preceding section, even quite modest trade costs can have a large impact. Furthermore, the non-monetary costs of international transactions are likely to be highly important for such activities. Synchronisation of production and the ability to respond rapidly to changing patterns of demand of costs may be important. “Just-in-time” production techniques may require very high-quality infrastructure, if not physical proximity1. Business travel, rapid port clearance and good electronic communications are likely to be important, as with the oft-quoted example of Intel’s decision to produce in Costa Rica. All these arguments suggest that these rapidly expanding export activities are deterred from investing in countries that are remote or that have poor infrastructure. The point of the example given in Figure 1 is that low wages may be insufficient to compensate for these trade costs. It is quite possible that wages cannot go low enough, as trade costs squeeze domestic value added to zero. 2.1.3
Trade costs and real income: absolute disadvantage
The wage schedule illustrated in Figure 1 is not the same as a real income schedule. To get to the real income penalty of trade costs, two further considerations have to be taken into account. One is that much of the economy produces non-tradable goods, and the other is the presence of an import competing sector. By non-tradable goods, we mean goods that will never be traded (haircuts). If these goods do not use any imported inputs, then the real income generated by these activities depends only on the productivity with which they are performed, regardless of anything happening in the traded goods sector of the economy. Thus, low wages in the economy translate into low prices for these goods, offsetting the real income loss. In contrast, goods in the import competing sector can either be produced domestically or imported. International trade allows the economy to import goods in which domestic production is less efficient; but trade costs pose an obstacle to this, causing the economy to forego some of the gains from trade. How large are the gains of specialisation foregone? This depends on the absolute disadvantage that the country has in its import competing sectors. A model for thinking about this is one in which there is a continuum of types of import competing product. Suppose that the world price of each of these goods is unity and that the efficiency with which the economy under consideration can produce them varies across the products – a variant of the famous Dornbusch, Fischer and Samuelson (1977) model of trade. More specifically, product types are indexed by z, and we denote the labour input required to produce good z in the country under consideration by a(z) so that if the wage in the economy is w, and (for simplicity) no other inputs are involved, domestic production costs are wa(z). The cost of importing the good is world price times the transport cost factor, pwt. In equilibrium, goods for which wa(z) < pwt are produced domestically, and goods for which wa(z) > pwt are imported. What is the effect of increasing trade costs? Higher trade costs reduce w, exactly as in the preceding subsection and in Figure 1, and raise t. Both effects operate to increase the set of goods that are produced domestically. The real income costs of this depend on the absolute disadvantage for the country of the goods that it is now producing domestically rather than importing. If this disadvantage is small, then trade costs will have a large effect on trade volumes, but little effect on real income. If the disadvantage is large, then imports will remain relatively large: it is extremely costly to replace imports with inefficient domestic production. Once again, we illustrate with an example. Figure 2 reproduces the domestic value added curve from Figure 1 (the solid line; and we now assume that labour is the only domestic primary factor). TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
- 65
The two dashed curves are real income, produced for different degrees of disadvantage in the import competing sector. In both cases the share of non-tradables in the economy is 50 per cent. Despite this high share, trade costs have a very substantial effect on real income. Along the upper dashed line, in which absolute disadvantage is relatively small – the volume of trade goes to zero at trade costs of around 1.5, and the economy has real income around 60 per cent of what it would have with free trade. Along the lower dashed line absolute disadvantage in import competing sectors is greater. Consequently, trade volumes are higher but real income lower, falling to less than half of what it would be with free trade.
Figure 2. Trade costs and real income
2.1.4
Infrastructure appraisal under perfect competition
While we have argued that trade costs can have a substantial effect on real income, the perfectly competitive framework that we have developed so far is consistent with standard approaches to infrastructure evaluations. There are no market failures, and standard cost-benefit analysis applies. Thus the gains are given by the direct cost saving on existing trade volumes, plus a part based on the value of the increase in trade, a “triangle” approximated by 0.5x(change in trade cost)x(change in trade volume). Of course, computing the change in trade volume may be difficult because, as we have seen, a non-marginal project has general equilibrium effects, changing specialisation and factor prices. The principles employed are, however, those of standard cost-benefit analysis.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
66 - INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE 2.2. Increasing returns to scale and intra-industry trade In the previous section, world prices of output were taken to be fixed, and production was assumed to be subject to constant returns to scale. For many products these are bad assumptions. Firms have increasing returns to scale, and so face the choice of where to locate a plant of minimum efficient scale. Products are differentiated and markets imperfectly competitive, so that prices are not taken as constant from the standpoint of a firm (and much less a country). In this sort of environment there is typically intra-industry trade: the product is produced by firms in a number of countries, each of which exports as well as selling to their local market. The questions are then: how do firms choose the countries in which to produce and, in particular, how do trade costs influence this choice?; and what do firms’ choices imply for the real incomes of more or less remote countries? 2.2.1
Market access
The standard approach for addressing this question is to look at a monopolistically competitive industry in which the number of firms producing in each country (or more generally, each “location”) is determined by profit opportunities. There are a number of possible locations, and the number of firms in the particular industry under study that produce in location i is denoted ni. These numbers are endogenous and adjust until profits are bid down to zero. Profits, and hence the equilibrium numbers of firms, depend on costs in each location and, if there are trade costs, on “market access”. The market access of location i is defined, loosely, as a weighted sum of demands in all possible markets (export as well as domestic), with weights depending inversely on the trade costs of supplying each market from location i. A precise definition of market access is given in annex, for the international trade variant of the Dixit-Stiglitz (1977) model of monopolistic competition. To see how market access shapes the location of firms, suppose that there are just two countries, i = 1, 2, with trade costs t > 1 between them and no internal trade costs within each country. Suppose also that country 1 is k > 1 times larger than country 2. Country 1 therefore has better market access than does country 2: a firm located in 1 can access more consumers at low trade costs. In this setup, where do firms locate? Or equivalently, what determines the value of n1/n2? If production costs are the same in both countries and there are trade costs between countries, it turns out to be the case that n1/n2 > k, i.e. production is more than proportionately skewed towards the location with good market access. The intuition can be seen by looking at a situation close to perfectly free trade. With free trade, n1/n2 = k is an equilibrium (although not the only one: if production costs are the same and there are no trade costs then firms are indifferent about where they produce). Adding small trade costs means that firms from the small country are disadvantaged in the larger market, and firms from the large country disadvantaged in the smaller market. Clearly, this makes the smaller country the less profitable location in which to produce: export sales, on which firms now bear trade costs, are a much larger part of activity for the firms in the small country than they are for firms in the large one. However, it is generally the case that some firms may continue to produce in the small country, as they are able to supply this, their local market, at relatively low cost. Production therefore moves towards the larger country, so that n1/n2 > k, although it may continue in both countries, implying that there is intra-industry trade despite the presence of trade costs. One further point needs to be made. If trade costs are extremely high – for example, autarky – then production has to take place in both countries in order to meet local demand, and we are back in the situation in which n1/n2 = k. This suggests that the impact of market access on location is greatest at trade costs that are in some sense “intermediate”: high enough that they influence firms’ location decisions, but not so high that they tie firms to producing just for the local market. TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
- 67
How does this story impact on wages? The larger economy has more labour than the smaller one, but production – and hence labour demand – is more than proportionately skewed towards the country with good market access. If wages are an increasing function of the quantity of labour demanded relative to that supplied, we therefore expect to see wages in location 1 higher than wages in location 2, w1/w2 > 1. This offsets but does not overturn the property that n1/n2 > k. In other words, the advantages of good market access show up partly in the economy having a relatively large number of firms operating in sectors where market access is important, and partly in the economy with good market access paying higher wages. This discussion is summarised in Figure 3, in which the horizontal axis is a measure of trade costs, as before, and the solid lines give nominal wages in the two countries w1 and w2 (see annex for details). The main point to note is that wages are higher in the economy with the better market access. Furthermore, wages are non-monotonic with respect to trade costs. It is at intermediate levels of trade costs that a country with poor market access is most disadvantaged. When trade costs are very low market access becomes unimportant, and when they are very high the small location benefits from natural protection; at intermediate levels its firms suffer more from import competition than they benefit from access to export markets. While the solid lines on the figure report nominal wages, the dashed ones give real wages (all scaled relative to their free trade values). The real wage adjustment increases the wage gap between countries of different sizes, essentially because the small country has to pay trade costs on a higher proportion of its consumption. It also means that the aggregate gains from reducing trade costs are greater – once again, because saving trade costs reduces the consumer price index.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
68 - INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
Figure 3. Trade costs and real income Country 1 larger than Country 2
Summarising, the main conclusions that come from Figure 3 are as follows. First, poor market access is associated with low real income, although the magnitude of the loss in real income is relatively low (compare the axes on Figures 2 and 3); of course, this would be magnified if the firms used imported intermediate goods. Second, it is possible (although not necessarily the case) that reducing trade costs brings real income losses; this possibility arises at relatively high trade costs for small countries, and lower trade costs for large countries. Third, the gains from reducing trade costs are largest for smaller countries gaining improved market access, e.g. for trade costs in the range 1-1.5 in the figure. In this range, reducing trade costs removes the disadvantage of peripherality. Finally, while we have illustrated the analysis with a two-country example, the insights extend to a multi-location world. It suggests that there is a real income gradient, declining from countries with good market access down to countries in the periphery. We look at the evidence for this in Chapter 3 below. 2.2.2
External economies of scale and agglomeration
The analysis outlined in the previous subsection was based on models in which economies of scale are internal to the firm. The role that economies of scale play in these models is to force each firm to choose to produce in just one (or a few) locations; if economies of scale were sufficiently small, then it would be profitable for each firm to produce in every location, in order to overcome trade costs. An additional possibility is that there are economies of scale that are external to the firm but spatially limited in their spread, so contained within a particular city, region or country. In this TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
- 69
case, the effects outlined in the previous subsection are amplified: the location with better market access has higher production levels, and hence more economies of scale, reinforcing its advantage and increasing the wage levels that, in equilibrium, it will pay. Agglomeration and clustering of activity may also occur. The location with good market access may attract all the firms in the sector, as the benefits of being in the cluster (receiving the external economies of scale from related firms) outweigh the possible costs of high input prices as wages and prices of other immobile factors are bid up. The sources of such external economies of scale are usually categorised into three types. First are technological externalities, for example, knowledge spillovers between firms in a particular area or sector – a Silicon valley effect. Second are benefits from a thick labour market. These can arise if there are industry-specific skills that can be acquired by a pool of workers; a large scale of activity may increase the incentives to acquire skills and reduce uncertainties, both about the value of acquiring specific skill and about the ease of finding suitably skilled workers. Third, there are pecuniary externalities deriving from supplier-customer relationships. The presence of trade costs means that firms benefit from locating close to their suppliers, and supplier firms will in turn benefit from locating close to customers. If these firms are imperfectly competitive then these input-output links create pecuniary externalities and increasing returns to the clustering of activity2. The implications of these sorts of external economies is drawn out in the “new economic geography” literature; for example, Fujita et al. (1999), who investigate the impact of agglomeration forces on the structure of production, trade and real incomes. For current purposes, there are two main things to note. The first is that the presence of these agglomeration effects amplifies the effects outlined in the previous section – the vertical axis on Figure 3 becomes stretched, as the advantages of market access determine the scale of activity in each location, and hence the magnitude of external scale economies. The second is that, if the external economies are large enough, then a cluster may form, drawing all activity into a single location. Such a cluster is – other things being equal – more likely to form in a location with good market access than bad, and its presence will create real income gains for the host location. Furthermore, once a cluster has formed it is likely to be persistent, and robust to small changes in parameters. Marginal changes in exogenous circumstances – such as infrastructure improvements – may not cause any relocation of activity. The lock-in of the existing cluster may be strong enough that no firm wants to leave, given the presence of other firms. 2.2.3
Infrastructure appraisal under imperfect competition
Market imperfections are an inherent part of the models of trade under imperfect competition and economic geography. Private marginal cost is below price, because of the market power of firms in the industry, and if there are positive externalities marginal cost is reduced still further. Expanding production in the industry is therefore socially valuable, bringing benefits over and above those that would be obtained by a standard narrow cost-benefit analysis. However, in this second-best world considerable care needs to be taken in drawing policy conclusions. For example, the appropriate marginal cost depends on whether output adjusts by changing output per firm or by changing the number of firms. Transport improvements may have a beneficial pro-competitive effect, as market integration increases the intensity of competition, and industrial reorganisation in response to this increased competition may bring further benefits of economies of scale. There are many theoretical ambiguities here, although simulation estimates in this type of model suggest gains from transport improvements that are 30-40 per cent larger than those given by narrow cost-benefit analysis (Gasiorek and Venables, 1998). This is an area where more work is needed.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
70 - INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE 2.3. Other mechanisms Before concluding the analytical chapter, we should of course note that there may be other mechanisms through which increased trade may increase real income. Coe and Helpman (1995) point to the role of trade in transferring the total factor productivity effects of R&D in one country to other countries, although Keller (2001) questions the extent to which trade is the mechanism through which such transfers occur. Better infrastructure and increased trade may also stimulate foreign direct investment (FDI). High trade costs can, in principle, either encourage or discourage FDI. The two possibilities correspond to the distinction between horizontal and vertical FDI. The former arises when firms’ investments are market oriented and involve duplicating downstream stages of the production process in order to serve the local market. For example, if trade costs are high then it may be more profitable to undertake production or assembly locally rather than import the final good. The latter, vertical FDI, corresponds to the case of “fragmentation” in which firms split their production process to locate segments in the minimum cost location, but the transport intensity of this activity makes it vulnerable to high trade costs (for a recent survey of theoretical and empirical material on FDI, see Barba Navaretti et al., 2004). The balance of evidence suggests that most FDI has been horizontal rather than vertical. However, there are several major qualifications. One is that this is essentially because most FDI has been between rich countries, with comparable factor costs; and even in this case good infrastructure is important. Horizontal investments may require a minimum scale of operation, in which case, access to a large “domestic” market is important; thus, market integration in the EU has encouraged horizontal FDI into export platform economies such as Ireland. The other qualification is that more recent evidence points to the growing importance of vertical FDI for developing and middle-income countries. As expected from our earlier discussion, this has gone particularly to countries that face relatively low trade costs, either because of proximity to large markets (Mexico, eastern Europe) or good infrastructure and trade links (Singapore and other parts of Asia). Finally, we note that FDI may be associated with technology transfer, although here too the evidence is mixed. Early studies found considerable evidence of spillovers but more recent studies, using more disaggregate data and better econometric techniques, have painted a much more mixed picture.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
- 71
3. TRADE AND REAL INCOME: EMPIRICS
With this review of some of the analytical issues in place, we now turn to look at the evidence. We structure the discussion in three sections. First, we look at issues involved in measuring the “remoteness” or market access of a country, then turn to evidence of its impact on the terms of trade, and finally to wage gradients, as suggested by theory. 3.1. Measuring market access Conceptually, there are (at least) three ways in which to measure trade costs and thereby determine a country’s remoteness or market access. One way is by direct measures, i.e. evidence on freight rates; a second is by using some proxy for freight rates, such as distance; and a third is by looking at actual trade flows, an endogenous variable. Our objective in this chapter is to assess the impact of these costs on real income, so we ideally need a single measure for each country, synthesising in an economically appropriate way the costs of trade in all goods, and with all actual and potential trading partners. A methodology for doing this, based on actual trade flows, was developed by Redding and Venables (2004) and in this section we outline the method, before using it to assess the costs of remoteness through the remainder of this chapter. Redding and Venables (2004) show how information derived from trade flows can be used to construct measures of the market access of countries, and demonstrate that the measures so constructed turn out to be exactly those which, according to theory, are appropriate determinants of the effects of trade on income. While the theory underlying this is given in the annex, the bare bones of the approach are as follows. Using international trade data, a gravity equation of trade flows can be estimated, based on the following equation:
Vij = s i T 1ij-σ mi . The left-hand side is the value of trade between countries i and j, and the right-hand side contains dummy variables for the export country, taking value si, and the importer country, mj . These estimates can be interpreted as measures of the supply capacity (si) of each exporter and the import demand potential (mj) of each importer. The term T 1ij-σ is an iso-elastic function of the cost of trade between the countries, and can be proxied by the usual measures – distance, common border etc. Estimating a log-linear version of this equation yields estimates for si, and mj and can be used to generate predicted values of T 1ij-σ . Using this information, two variables, which Redding and Venables call market access (MA) and supplier access (SA), can be constructed for each country:
MAi = Σ j T 1ij-σ m j . SAi = Σ j s j T 1ij-σ . The former is the sum of the effective import potential of each country, mj, from the trade equations, weighted by a measure of its effective distance from country i, where this measure is the TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
72 - INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE economically relevant one as estimated from the bilateral trade equations. This measure is conceptually similar to the old economic geography concept of “market potential”. A typical measure of the market potential of country i might be the sum of all countries’ incomes, each weighted by their distance from country i. The measure of market access given above, MAi, replaces income by a measure of imports, and distance by a predicted value, both derived from econometric analysis of trade flows. The second measure, supplier access, is the analogous concept, but measuring not each country’s access to markets but its access to sources of supply, important for thinking about the costs of intermediate and capital goods imports. It is formed as the sum of each country’s effective export supply, sj, once again appropriately weighted according to the “distance” to country i. These two measures turn out to be highly collinear, but nevertheless provide economically meaningful ways of measuring countries’ access both to markets and to sources of supply. One further step has to be noted, however. While the trade data reveals values of the importance of foreign markets, it does not reveal the importance of the home market – there is no estimate of T 1ii-σ . In what follows we shall refer to the foreign parts of market access and supplier access as FMA and FSA respectively, and discuss the way in which adjustments can be made to include the home market effect. 3.2. Supplier access and prices Armed with these measures, we first look at the extent to which remoteness raises the prices of imports. We focus on the relative price of Machinery and Equipment, since we want to identify the extent to which value added in production is reduced by having to pay high prices for inputs. Data on the relative price of Machinery and Equipment is available from the United Nations International Comparisons (ICP) project (United Nations, 1994), which gives information on the price of a large number of individual commodities in local currency units per dollar. Our measure of the relative price of Machinery and Equipment is the PPP for Machinery and Equipment divided by the PPP for GDP as a whole. Data are available for 46 countries for the year 1985. The relative price of Machinery and Equipment is 1 in the United States and reaches a maximum of 4.68 in Sri Lanka. Figure 4 presents a plot of this price measure against FSA, our measure of how far countries are from sources of export supply. We see a clear and statistically significant negative relationship, with countries with good market access having relatively cheap equipment. The effects are large, with New Zealand, Ethiopia and Zambia having relative prices of machinery and capital goods around 65 per cent higher than in the US (e0.5 = 1.65), while Kenya, Thailand and Senegal have prices around twice as high. Eaton and Kortum (2000) also point to the role of distance in impeding trade, particularly in capital goods and equipment, and draw out the implications of this for investment decisions and hence for growth. 3.3. Market access and real income The same analytical framework that generates the market access and supplier access measures also suggests that real income in each country should be a log-linear function of these two measures. As we have seen, the two measures are collinear, so for present purposes we concentrate on the relationship between real income and market access alone.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
Figure 4. Relative price of machinery and equipment and FSA
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
- 73
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
Figure 5. GDP per capita and FMA
74 - INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
Figure 6. GDP per capita and MA = DMA + FMA
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
- 75
76 - INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE Figure 5 presents the scatter-plot of the relationship between FMA and per capita income (both measured in logs, country codes given in the annex), illustrating a strong positive relationship between the variables. For example, looking just within Europe, there is evidence of a wage gradient from Belgium/Luxembourg (countries with the best foreign market access) through France, Britain, to Spain, Portugal and Greece. Several other points stand out. One is that a number of countries are able to escape the consequence of remoteness from export markets – e.g. Australia, New Zealand, Japan, the USA. However, looking at the bottom right-hand area of the figure, good foreign market access provides a safety net against very low incomes – despite the relatively poor performance of former communist countries. Market access is derived both from proximity to export markets, and from access to a large domestic market. Both proximity and scale matter. The scale effect is absent from Figure 5, but is included in Figure 6, where the horizontal axis is the sum of foreign market access (FMA) and “domestic market access” (DMA) a measure of domestic market size adjusted for the area of the country. Combining these effects provides very strong evidence of a wage gradient, indicating the importance of both proximity and scale in determining income levels. The impression given by these scatter-plots stands up to econometric analysis. Both FMA alone and the full MA measure are robust and statistically significant determinants of GDP per capita under a number of different econometric specifications. Controls are added for natural resource abundance, for pure geographical effects (distance from the equator, prevalence of malaria), for political and institutional factors (expropriation risk, formerly communist and civil war variables) and for regional effects. Potential endogeneity issues are addressed by the use of instrumental variables. The elasticity of per capita income with respect to FMA is robust, and approximately equal to 0.25 (see Redding and Venables, 2004, for details). One way of illustrating the quantitative importance of market access is to undertake a set of hypothetical experiments of the form: suppose we move country 1 to the location of country 2 then, holding other things equal (its institutional quality and resource endowments), what would happen to country 1's income? Tables 1 and 2 report the results of a few experiments of this type. Being landlocked and being an island both have a negative effect on real income, and the first column of Table 1 indicates that the penalty for being landlocked is substantial – removing it would raise income by one-quarter3. The cost of island status is smaller, costing around 7 per cent of GDP [column (2)]. Column (3) reports a trade policy experiment: changing countries’ trade openness [as measured by the Sachs-Warner (1995) openness index] from the 1994 value to the most open possible. This too yields extremely large income gains, of around 25 per cent for countries that were, in 1994, quite economically closed. Column (4) reports the experiment of moving a country from its present location to that of Hungary, on the edge of the EU. The dramatic increase in FMA brought about by this change means that for some of the most remote economies in the sample income increases by nearly 80 per cent.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
- 77
Table 1. Percentage change in real income from changes in market access Variable
Country
(1)
(2)
(3)
(4)
Access to coast
Loss of island status
Become open
Distance (move to Central Europe)
Australia
7.3%
Sri Lanka
7.3%
20.7%
67.4%
Zimbabwe
24.0%
27.7%
Paraguay
24.0%
25.3%
79.7%
Hungary
24.0%
26.5%
58.3%
Notes: Source:
Actual values for the Sachs and Warner (1995) openness index are 1 in Australia, 0.2321 in Sri Lanka, 0.038 in Hungary, 0.077 in Paraguay, and 0 in Zimbabwe. Redding and Venables (2004).
Common borders are also important for facilitating trade and improving market access, and Table 2 quantifies their importance by the hypothetical experiments of closing borders. The effects reported show that smaller countries gain very substantially from access to a large neighbour, as illustrated for Mexico and the Czech Republic. However, two small neighbours, the two African economies, neither of which has large markets nor supplies of manufactures to offer, only experience extremely small border effects. An implication of this is that South-South regional integration schemes yield very limited benefits compared to fuller integration into the world economy as a whole.
Table 2. Percentage change in real income from border effects Removal of common border Germany-Czech Republic
US-Mexico
Zimbabwe-Zambia
Effect on per capita income Germany
Czech Republic
- 0.1%
- 25.7%
US
Mexico
- 0.5%
- 27.2%
Zimbabwe
Zambia
- 0.05%
- 0.11%
Source: Redding and Venables (2004).
3.4. Market access and income: conclusions The empirical results of this chapter are broadly consistent with theory, suggesting large real income gradients from location with good market access to the periphery. We have presented TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
78 - INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE empirical results just from one framework – based on monopolistic competition and market access – but note that the effects of infrastructure and trade can also be identified through other approaches. For example, Gallup et al. (1998) emphasize the combined effects of distance and malaria: Eaton and Kortum (2000) put emphasis on prices of capital goods and consequent disincentives to invest; Martins and Winters (2004) highlight both distance and smallness. Of course, there is a larger literature on other determinants of per capita income (and its growth), but our focus here is on the transmission mechanism through trade costs and market access.
4. TRADE COSTS, TRADE FLOWS AND INFRASTRUCTURE
The preceding chapters showed how per capita income is determined by trade costs and by market access. The market access measures that we used were themselves constructed from observed trade flows and from distance. The final set of questions goes a stage further back, and asks: what determines trade costs and trade flows, and in particular what is the role of infrastructure? The determinants of trade volumes and of trade costs, and the nature of the relationship between them, has been the subject of a large literature, notably by Hummels (2001 and other papers) and Anderson and van Wincoop (2004), and this chapter will not undertake a full review of this literature. Instead it will present some selected findings, focusing on the role of infrastructure. Combining estimates from different sources enables us to derive some crude estimates of the importance of infrastructure, via its impact on trade, for real income. 4.1. Determinants of trade costs The first column of Table 3 reports estimates, drawn from Limao and Venables (2001), of the determinants of trade costs. The dependent variable is the cif/fob ratio for bilateral trade flows between 103 countries. This variable is far from ideal, failing to capture many elements of trade costs and in some cases using imputed values. (See Limao and Venables, 2001, for discussion of some of the issues and presentation of similar results based on actual freight charges, as reported by shipping companies.) The independent variables fall into three categories. First are the geographical ones; distance between countries, and dummy variables for whether countries share a common border and for whether either country is an island. (In the table, a prefix p on a variable denotes the importer’s trading partner in the bilateral relationship.) Distance and border effects are significant with the expected signs, and the fairly low elasticity of trade costs with respect to distance – a 10 per cent increase in distance is only associated with a 2.1 per cent increase in trade costs – reflects the importance of loading and port costs relative to pure distance. The second set of variables shows per capita incomes, and these have negative impact, possibly reflecting low productivity and poor transport services. The third set of variables attempts to draw out infrastructure effects more explicitly. The infrastructure measure used is designed to gauge the costs of travel in and through a country. It is constructed as an average of the density of the road network, the paved road network, the rail network and the number of telephone main lines per person. The regressions use an inverse measure of this index, so that an increase in the variable inf is expected to be associated with an increase in the costs TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
- 79
of transport. We see that own and partner country infrastructure (inf and pinf, respectively) have significant effects on trade costs, with the expected sign. The final two variables take non-zero values only for land-locked countries, and give the impact of infrastructure in the transit country through which surface trade has to pass; again, the effects are significant and of the expected sign. We discuss the economic magnitude of these effects in section 4.3 below. Table 3. Determinants of import volumes and importers trade costs (bilateral, country and partner country) Dependent variable: ln Bilateral trade costs
0.21*** (5.65) -1.36*** (-7.78)
Dependent variable: ln Bilateral imports 0.99*** (28.04) 1.28*** (34.67) -1.37*** (-18.03) 2.52*** (7.25)
-0.09 (-1.23) -0.12 (-1.65) -0.23*** (-9.64) -0.30*** (-12.84) 0.34*** (3.92) 0.66*** (7.64) 0.36*** (2.15) 0.24*** (2.51)
0.35** (2.46) 0.4*** (2.78) 0.16*** (2.96) 0.16*** (3.04) -1.32*** (-7.49) -1.11*** (-6.26) -0.6*** (-3.04) -0.45** (-2.26)
0.48 4615
0.8 4516
lnY lnpY Ln Distance border isldummy pisldummy lnY/cap lnpY/cap lnInf lnpInf ln(1+Inftran) ln(1+pInftran)
Pseudo Rsq No. of observations
Notes: 1) N=4516; Tobit estimates. Pseudo Rsq given by the correlation of actual and predicted values; constants included but not reported; σ, the standard error of the Tobit estimate. 2) T-statistics in parenthesis ; ***, **,* indicates significance at 1 per cent, 5 per cent and 10 per cent levels, respectively. Constant included in all specifications but not reported. Source: Limao and Venables (2001).
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
80 - INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE 4.2. Determinants of trade flows The standard tool for estimating the determinants of trade flows is the gravity equation, which relates bilateral trade flows to country characteristics, such as income, and to between-country characteristics, such as distance or whether a pair of countries share a common border, language or currency. The effects of distance are large, typically giving an elasticity of trade flows with respect to distance of around -1, so doubling distance halves trade flows. Column 2 of Table 3 gives results from Limao and Venables (2001). The dependent variable is now bilateral trade between a pair of countries, and the first two independent variables scale this to country size: as would be expected, there are coefficients of around unity on both importer and exporter GDP. Other variables are consistent with the determinants of trade costs, and are all significant (whereas the effect of island status in reducing trade costs was insignificant, its effect on raising trade volumes is significant). Once again, the infrastructure variables come through strongly. It is natural to try and combine results from the two exercises reported in Table 3 and ask, what is the impact of trade costs on trade volumes? Limao and Venables (2001) discuss alternative ways in which this effect can be estimated, and come up with a central estimate of the elasticity of trade volumes with respect to the trade cost factor of around -3. Thus, doubling trade costs from their median value, i.e. raising the trade cost factor from 1.28 to 1.56, reduces trade volumes by 45 per cent. 4.3. The role of infrastructure The results given in Table 3 point to the importance of infrastructure in determining both trade costs and trade volumes. To quantify the magnitude of the infrastructure effects, we use the estimated elasticities given in Table 3 to establish the costs and benefits of being at different points in the distribution of countries by infrastructure. We take the country with median infrastructure (which turns out, in this large sample of countries, to be Pakistan), and ask what would happen if its infrastructure were improved up to the quality of the country at the 25th percentile (the US) or deteriorated to the level of the country at the 75th percentile (Nigeria), holding other independent variables constant. Results are given in Table 4. Thus, improving infrastructure from the median to the 25th percentile, holding other things constant, reduces the cif/fob ratio from its median value of 1.28 to 1.12; it raises the volume of trade by a full 68 per cent. Going in the other direction, a reduction in infrastructure quality to the 75th percentile raises the cif/fob ratio to 1.39 and reduces trade volumes by 28 per cent.
Table 4. The role of infrastructure
Trade costs (cif/fob) Trade volumes relative to median Real income
Infrastructure at 25th percentile 1.12
Median infrastructure 1.28
Infrastructure at 75th percentile 1.39
1.68 1.14
1.00 1.00
0.72 0.92
Source: Based on results in Limao and Venables (2001).
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
- 81
One final and even more speculative step can be taken. If we combine these estimates with results from Chapter 3, how much does changing infrastructure change income by? We suggested in section 3.3 that the elasticity of GDP per capita with respect to FMA was of the order of 0.25. Since FMA is proportional to trade volumes, we can combine the elasticity of income with respect to FMA and the elasticity of trade volumes (hence FMA) with respect to infrastructure. Effects are reported in the final row of Table 4. This suggests that improving infrastructure from the median to that of a country at the 25th percentile would – through the mechanism of foreign trade alone – raise per capita income by some 14 per cent. Deteriorating infrastructure to the 75th percentile would reduce income by 8 per cent.
5. CONCLUDING COMMENTS
The results surveyed in this paper indicate that infrastructure is a significant determinant of trade costs and that trade costs – and derived concepts such as market access – are important determinants of per capita income levels. The mechanisms for this include direct terms-of-trade effects, impacts on the extent to which countries are able to specialise in line with comparative advantage, and on the incentives for firms to operate in different locations. It is likely that results discussed here are lower bounds; transport infrastructure is important not just in determining direct shipping costs, but also in determining time in transit and the reliability of trade. The impact of infrastructure may be felt not just through increased trade flows but also through enhanced face-to-face contacts, knowledge spillovers and technology transfer. There are many areas where further research is needed. The high monetary value of time saved has been demonstrated by Hummels (2001b), but puzzles remain. Exactly why are just-in-time management processes valuable, and how dependent are they on physical proximity? How difficult is it to monitor or control a remote operation, and how does infrastructure mitigate these problems by facilitating business travel and face-to-face contact? Integrating the economic analysis of these issues with management approaches is important if we are to better understand the full economic impact of logistics systems and alternative management practices. Transport appraisal techniques need accurate measures of the direct benefits of infrastructure improvements, and also need to take into account the full equilibrium implications of these benefits. How do improvements affect the location of activity, pattern of specialisation and factor prices in a country? What are the market failures that need to be taken into account in cost-benefit analysis? Empirical analysis is difficult, because it is hard to identify the full equilibrium implications of particular infrastructure projects, yet these effects are crucial in establishing the full benefits of such projects.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
82 - INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
NOTES
1.
See Harrigan and Venables (2003) for an argument that delivery delay is qualitatively different from monetary trade costs, tending to promote clustering of activity.
2.
For discussion of these agglomeration forces, see Henderson and Thisse (2004).
3.
The model specification means that the same proportional effect is experienced by all countries.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
- 83
ANNEX
Chapter 2, Figures 1, 2: Economy endowed with one unit of labour with wage rate w, so national income is w.
(
)
Utility: u = w / wθ G 1−θ . . The denominator is the consumer price index defined on the price of the non-traded good (= w) and the price index of the import competing sector, G. Expenditure shares of non-traded and import competing goods are respectively 2, 1 - 2. Export sector: Fob price 1/t: Price = unit cost: 1/t = c(w, r, t) Import competing sector: Import price t: produce goods with index z < z*, wa(z*) = t. z*
Price index, G = t( 1 − z*) +
∫ wa( z )dz 0
These four equations solve for equilibrium u, G, w, z*. In the example of the figure,
a(z) = exp[∀+∃z].
c(w, r, t) = [∀ww(1-Φ) +∀rr(1-Φ) + ∀mt(1-Φ)](1/(1-Φ)) Figure 1, Φ = 0.67: top line: ∀w = 0.5, ∀r = 0, ∀m = 0.5: Lower line, ∀w = 0.25, ∀r = 0.25, ∀m = 0.5. Figure 2: 2= 0.5; ∀w = 0.5, ∀r = 0, ∀m = 0.5; ∀ =0, ∃ =2, ∃= 5.
Chapter 3, Figure 4: For full reference on the Dixit-Stiglitz monopolistic competition model, see Fujita et al. (1999). The size of each economy is proportional to parameter E1 = 3, E2 = 1. Figure 3 is produced using wage equation,
[
σ wi = ∑ j E j G j -1 T 1ij-σ
]
1/σ
[
with price index G i = ∑ j n j w1j−σ T 1ij-σ
]
1 /( 1−σ )
, Φ = 5.
Labour market clearing gives a further equation linking labour demand (proportional to the number of firms) to the wage rate; we assume that this takes the form ni = Ejwi 0.5. These three equations solve for equilibrium values wi, ni , Gi with k = E1/E2 = 3. The real wage is wi/Gi0.25 (i.e. this sector accounts for 25 per cent of consumption).
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
84 - INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
BIBLIOGRAPHY
Anderson J. and E. van Wincoop (2004), “Trade Costs”, Journal of Economic Literature, 42, 3, September. Barba Navaretti, G. et al. (2004), Multinational firms in the World Economy, Princeton University Press. Coe, D. and E. Helpman (1995), “International R&D spillovers”, European Economic Review, 39, 859-887. Dixit, A.K. and J.E. Stiglitz (1977), “Monopolistic competition and optimum product diversity”, American Economic Review, 67, 297-308. Dornbusch, R., S. Fischer and P.A. Samuelson (1977), “Comparative advantage, trade and payments in a Ricardian model with a continuum of goods”, American Economic Review, 67, 823-839. Eaton J. and S. Kortum (2001), “Trade in capital goods”, European Economic Review, Elsevier, Vol. 45(7), 1195-1235. Evans, C. and J. Harrigan (2005), “Distance, Time, and Specialization: Lean Retailing in General Equilibrium”, American Economic Review, Vol. 95(1), 292-313(22), March. Frankel, J. and D. Romer (1999), “Does trade cause growth?”, American Economic Review, 89(3), 379-399. Fujita, M., P. Krugman and A.J. Venables (1999), The Spatial Economy: Cities, Regions, and International Trade, MIT Press. Gallup, J., J. Sachs and A. Mellinger (1998), Geography and Economic Development. Proceedings of World Bank Annual Conference on Development Economics. World Bank, Washington. Gasiorek, M. and A.J. Venables (1998), The Welfare Implications of Transport Improvements in the Presence of Market Failure, report to SACTRA, UK Department of Transport. Hanson, G. (1998), “Market potential, increasing returns and geographic concentration”, NBER Working Paper, 6429. Harrigan, J. and A.J. Venables (2004), “Timeliness and agglomeration”, CEPR discussion paper, 4294. Harris, C. (1954), “The market as a factor in the localization of industry in the United States”, Annals of the Association of American Geographers, 64, 315-348.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
INFRASTRUCTURE, TRADE COSTS AND THE GAINS FROM INTERNATIONAL TRADE
- 85
Helpman, E. and P. Krugman (1985), Market structure and foreign trade, Cambridge, MIT Press. Henderson, V. and J. Thisse (2004), (eds.) Handbook of Urban and Regional Economics, Vol. 4, North-Holland, Amsterdam. Hummels, D. (2001), “Towards a geography of trade costs”, Purdue University, mimeograph. Hummels, D. (2001b), “Time as a trade barrier”, Purdue University, mimeograph. Keller, W. (2001), “International technology transfer”, NBER working paper 8573. Krugman, P. (1991), “Increasing Returns and Economic Geography”, Journal of Political Economy, 99(3), 483-99. Krugman, P and Venables, A.J. (1995) “Globalisation and the Inequality of Nations”, Quarterly Journal of Economics, 110(4), 857-80. Limao, N. and A.J. Venables (2001), “Infrastructure, geographical disadvantage, transport costs, and trade”, World Bank Economic Review 15, 451-479. Overman, H.G., S. Redding and A.J. Venables (2003), “The Economic Geography of Trade, Production and Income: A Survey of Empirics”, in: Harrigan, J. (ed.), Handbook of International Trade, Basil Blackwell. Redding, S. and A.J. Venables (2004), “Economic geography and international inequality”, Journal of International Economics, 62, 53-82. Sachs, J. and A. Warner (1995), “Economic reform and the process of global integration”, Brookings Papers on Economic Activity 0(1), 1. Venables, A.J. and N. Limao (2002), “Geographical disadvantage: a Heckscher-Ohlin-von Thunen model of international specialisation”, Journal of International Economics, 58, 239-263.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
SUMMARY OF DISCUSSIONS -
SUMMARY OF DISCUSSIONS
Andreas KOPP Chief Economist OECD/ECMT Transport Research Centre
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
87
SUMMARY OF DISCUSSIONS -
89
SUMMARY
1.
INTRODUCTION ......................................................................................................................... 91
2.
TRANSPORT COSTS AND THEIR IMPACT ON INTERNATIONAL TRADE...................... 92
3.
THE DEVELOPMENT OF INTERNATIONAL TRANSPORT COSTS.................................... 94
4.
TRANSPORT COSTS, INTERNATIONAL TRADE AND ECONOMIC DEVELOPMENT ... 96
5.
TRADE FACILITATION ............................................................................................................. 98
BIBLIOGRAPHY ............................................................................................................................... 101
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
SUMMARY OF DISCUSSIONS -
91
1. INTRODUCTION
The Round Table started out from a discussion of how international transport costs have developed in the last decades. In contrast to the popular assumption that the currently observed globalisation is to a large extent due to the decrease of transportation costs, the empirics show a more differentiated picture: –
The traditional measure of transport costs per tonne, which is important for estimates of the development of transport equipment use and infrastructure planning, does not show a secular, overall decline in international transport costs. Ocean shipping prices increased from the beginning of the seventies until the mid-eighties and have declined since then. Prices of maritime transport services per tonne are, however, about the same as they were at the beginning of the seventies.
–
International transport costs per tonne have declined for air transport, although not in such a dramatic way that it could account for the strong intensification of international trade.
–
More important for the measurement of the trade frictions resulting from transport costs is the cost of transport service per value of goods traded. While corresponding prices for ocean shipping have decreased for some countries and some goods, this is not a universal phenomenon.
–
Transport costs per value unit have substantially decreased for air transport.
–
The increase in international trade has been accompanied by a massive increase in the value/weight ratio of the goods traded. This implies the increased impact of variations in international transport costs on the incentives to trade goods and services internationally.
–
The structural shift of trade flows has been accompanied by a strong shift in the modal split of international transport.
A first indication of the importance of transport costs and transport policies for international trade is obtained from empirical studies of the importance of distance for international trade relations. Roughly half of the world’s trade takes place between countries located within 3 000 kilometres of each other. This picture has not been changed significantly by the fact that long-distance transport costs decrease more sharply than costs for short-distance transport. Bilateral trade costs are strongly influenced by trade volumes, indicating that trade costs are dependent on the fixed costs of transport infrastructure facilities. For smaller countries, this implies high benefits from co-ordinated investment in international transport infrastructure. The Round Table discussed the consequences which transport costs have for national incomes if the economies export part of their production and import part of their intermediate goods. The poorer the countries are, the more likely it is that the income costs of high trade costs will have to be borne by the wage earners. If the production exhibits increasing returns to scale, transport costs also strongly TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
92 – SUMMARY OF DISCUSSIONS influence firms’ location decisions. In general, the accessibility and size of an economy favour location decisions by monopolistically competitive industries. Small countries could suffer from the reduction in transport costs if the costs remain in an intermediate range. Transport policy has a particular role to play in reducing transport costs by infrastructure investment. The background papers and the discussion at the Round Table confirmed that infrastructure investment plays a crucial role for the level of trade costs. As the benefits of transport infrastructure investment, which helps international trade, do not only accrue to the domestic population, there is a strong need for international co-ordination in order to avoid a persistent tendency towards underinvestment in international transport infrastructure. When policy barriers to international trade are reduced, behind-border policies, traditionally considered to be national policies, gain importance to help further international economic integration. The Round Table discussed four areas of reform which bear upon international trade costs: measures to increase “port efficiency” (broadly defined); the reform of customs procedures; investment in IT infrastructure; and regulatory reform. The background papers to the Round Table discussion showed large trade gains to be expected from these reform steps globally, and in Middle and Eastern European countries in particular. The analysis revealed major differences in distances from best practice benchmarks between the new EU-8 member countries and the sample candidate EU member countries. With respect to a possible sequencing of measures, it is important to note that the greatest absolute trade gains, of US $49 billion and US $62 billion, would result if the port efficiency and IT infrastructure reached half the average level of the EU-15. Seventy per cent of these gains would result from export expansion. Again, initiatives to facilitate trade are associated with international co-ordination problems, as the benefits of their implementation do not only occur domestically but also in the trading partners’ countries. Without co-ordination mechanisms, it is likely that too little effort will be allocated to trade facilitation measures.
2. TRANSPORT COSTS AND THEIR IMPACT ON INTERNATIONAL TRADE
There is widespread agreement that the reduction in long-distance transport and communications costs has been an important determinant of today’s globalisation. In view of this consensus, relatively little research has been carried out on what transport policy’s role has been in bringing about the decline in transport costs and, more importantly, which challenges transport policy has to face in a globalising world and how to respond to them. One reason for this contrast has been the fact that transport costs, or more generally trade costs other than trade policy barriers, have traditionally been largely neglected, not only in international economics but in macroeconomics more generally. Obstfeld and Rogoff (2000) argue, for example, that all the major unresolved problems of international macroeconomics hang on trade costs.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
SUMMARY OF DISCUSSIONS -
93
Trade costs, broadly defined, include not only transport costs but all costs incurred in getting a good to the final user, minus the marginal costs of production: policy barriers (tariffs and non-tariff barriers), information costs, contract enforcement costs, costs associated with the use of different currencies, legal and regulatory costs and local distribution costs (wholesale and retail). Transport costs are not only freight but also time costs. Given the broad definition of trade costs, it is clear that transport policy plays a central role in their determination rather than only in narrowly-defined transport costs. For a long time it was believed that trade costs other than policy barriers would be of minor importance for the volume and structure of international trade. Recently, it has been acknowledged that trade costs are large and varied. A prominent example is given in the literature (Feenstra, 1998) to illustrate the potentially large effects: the production costs for Mattel’s Barbie doll is 1 US dollar, while it sells for 10 US dollars on the US market, without any policy barrier to its importation. The costs of transportation, marketing, wholesaling and retailing represent an ad valorem tax equivalent of 900 per cent. A rough overall estimate of the tax equivalent of representative trade costs for industrialised countries is 170 per cent. This breaks down into 21 per cent transport costs, 44 per cent border-related trade barriers and 55 per cent retail and wholesale distribution costs (1.7 = 1.21 x 1.44 x 1.55 - 1). The 21 per cent transport costs estimate includes directly measured freight costs and a 9 per cent tax equivalent of the time value for goods in transit (Anderson and Wincoop, 2004, p. 692). By contrast, direct evidence on border costs shows that tariff barriers are now low, on average less than 5 per cent for rich countries and, with a few exceptions, average between 10 and 20 per cent for developing countries. High value-to-weight goods are less penalised by transport costs. The value of the timeliness of transportation varies across goods and over time, explaining the modal split and its gradual development. The empirical picture of how transport costs influenced and influence international trade was presented by David Hummels in his paper, Transportation Costs over Time. The impact of transport costs on international trade points to the importance of transport policy for the overall economic development of national economies, and depends on their access to export and import markets. The paper by Anthony Venables, Infrastructure, Trade Costs and Gains from International Trade, introduced the mechanisms by which transport policy can help to reduce trade costs and how reduced trade costs change the real income of the countries involved. Although distance is a powerful explanatory variable for international trade, trade theory did have relatively little to say about its implications. The question of how transport costs determine not just the volume of trade but industrial structure, factor prices and income across countries is particularly important for low-income countries, some of which, even after trade liberalisation, find participation in world trade impeded by transport costs and other real trade barriers. Long distances to relevant export and import markets, having poor infrastructure and/or being landlocked by neighbours with inadequate infrastructure can make transport costs many times higher for some poorer countries than for developed countries. For these countries, trade costs have dramatic effects in reducing trade volumes. Estimates suggest that doubling transport costs approximately halves trade flows, and landlocked countries have 50 per cent higher transport costs than otherwise similar coastal economies (Venables and Limão, 2002). Gallup and Sachs (1999) and Radelet and Sachs (1999) have shown that high transport costs can damage countries’ export performances and growth. The Round Table discussed in detail the role of transport policy in reducing international trade costs, which crucially depend on the structure of product markets. Consequently, the planning and TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
94 – SUMMARY OF DISCUSSIONS evaluation of transport policy measures have to be based on an analysis of the sectors demanding transport service, including export and import sectors. Within the category of transport policies which help international trade, trade facilitation measures are of particular importance. The background paper of Wilson, Luo and Broadman, Trade and Transport Facilitation: European Accession and Capacity Building Priorities, examines trade facilitation and capacity building priorities in twelve countries in Europe and the Central Asia region. Based on data from the World Economic Forum, indicators of port efficiency, regulatory regimes, IT infrastructure and customs regimes were analysed to determine the most effective instruments to reduce trade costs. The Round Table discussed the reform steps, considered in the light of the differences between individual countries and the progress achieved so far. Measures to facilitate trade promise substantial benefits for the countries which are part of the integration process.
3. THE DEVELOPMENT OF INTERNATIONAL TRANSPORT COSTS
Direct international transport costs include freight charges and insurance, which is customarily added to the freight charge. Indirect transport costs include holding costs for the goods in transit, inventory costs due to buffering the variability of delivery dates, preparation costs associated with shipment size, etc. There is usually no direct evidence of the indirect cost of transport. Direct information on transport costs is obtained by quotes from shipping firms for standard transport services (as in Hummels, 2001b), on ocean shipping and air freight. Indirect sources are US census data on imports – by exporter country, mode of transport and entry point, by weight, and valued at cif and fob prices. A more widely available but less satisfactory source of ad valorem transport costs are the aggregate bilateral cif/fob ratios produced by the IMF from matching export data (reported in fob values with import data reported in cif values)1. A first indication of the development of international transport costs can be obtained by looking at the import wedge, i.e. the price of goods at the exporter’s departure port relative to the price at the importer’s destination. The data on the development of the import wedge presented to the Round Table came from customs declaration forms, in which firms report the cif and fob values of the shipment measured. In contrast to widespread perceptions about what has driven the recent increase in international trade, the ad valorem shipping costs for all goods have not decreased for most of the countries in the sample and between 1994 and 2000. Shipping costs create a substantial wedge between the domestic and foreign prices, except for the US whose imports are dominated by North American goods with very low shipping rates. Rates differ substantially across products, with ad valorem costs being much higher for bulk commodities than for manufactured goods. The aggregate import wedge has not decreased much either. Data for New Zealand and the US, countries for which longer time series are available, show that ad valorem freight costs have been fairly stable on average for New Zealand over the last four decades, and have decreased for the US since the time of the first oil shock. However, behind the stable aggregate figures lie major shifts from ocean shipping to air transport and dramatic differences in the development of shipping costs across modes. Leaving aside very large price spikes in the oil shock years, time charter series for ocean shipping show no clear decline, while TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
SUMMARY OF DISCUSSIONS -
95
the voyage charter series exhibit a downward trend relative to the US GDP deflator. Also, the liner shipping price index of the German Ministry of Transport shows a strong upward trend from 1954 to 1984 and a downward trend since then, with liner shipping prices in 2004 being about the same as in 1970. A clear downward trend in transport costs is observed for air transport. Measured in prices per kilogram shipped, there is a worldwide average of 1.5 per cent per annum. Declines in ad valorem rates are much larger. There is a decrease of 3.5 per cent over all routes. The sourcing wedge compares prices for two different foreign sources of supply. The size of the sourcing wedge is mainly influenced by distance and transport scale economies. In contrast to what have been considered to be the most important determinants of trade flows in the theoretic literature, distance has been and remains a crucial factor to explain international trade. Roughly half of world trade takes place between countries located within 3 000 kilometres of each other. While the increase in transport costs according to distance decreases over time, this had remarkably little influence on the impact of distance on trade. In fact, there is an ongoing debate on whether the results on the importance of distance for trade should not give reason to doubt the usefulness of the methods of analysis. For example, Grossman (1998) pointed out that the distance coefficients found in empirical studies implied that regions which are 500 miles apart trade 2.67 times more with each other than regions that are 1 000 miles apart. Introducing information barriers (Portes and Rey, 2002) or levels, instead of logarithmic argument values in the estimation equations, reduces the coefficient value for distance (Coe et al., 2002) without removing the strong influence of distance; indicating in turn the strong influence of transport costs and transport policy on international trade. The fact that the influence of distance on international trade has not decreased in recent years has been termed the “missing globalisation puzzle”. It is based on the finding in a large number of studies that the distance elasticity has not declined or has even risen over time (see a survey of this literature in Coe et al., 2002). The perception of a puzzle is due to the presumption that transport costs have generally declined. Attempts to solve the puzzle have been inconclusive so far (e.g. Brun et al., 2002). A second major factor in the size of the sourcing wedge is scale economies, or rather decreasing average costs in shipping. The evidence shows that larger importers have clearly smaller shipping costs for comparable goods. More systematically, Hummels and Skiba (2004) have estimated that doubling bilateral trade quantities results in a 12 per cent reduction in shipping costs. One of the reasons for decreasing transport costs in international trade lies in the indivisibility of transport infrastructure. The high fixed costs, for example for port facilities, point to an important demand for international transport policy co-ordination between smaller countries. The Round Table discussed the observed co-ordination failures in Europe and elsewhere and the ensuing disadvantage of high international trade costs. Other cost advantages correlated with trade volumes result from an increasing set of specialised services, as well as the profitability of larger vessels and aircraft with lower unit costs for transportation services. The link between transport costs and international trade has dramatically changed in character, due to a shift in the composition of world trade with respect to value per ton over the last thirty years. WTO data show that the real value of trade grew 18-fold in real terms between 1970 and 1999, while trade in manufactures grew 22-fold, trade in agricultural and mining goods grew only to a ten-fold level. With these structural shifts, the value/weight ratio has increased by more than 900 per cent since 1970.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
96 – SUMMARY OF DISCUSSIONS The increase in the value/weight ratio is mirrored by the change in modal shift. Air shipment is more likely to be used when the ad valorem price differential between the modes is small. This is the more so the smaller the freight bill relative to the unit value of the good, that is, when the value/weight ratio is large. One reason for the increase in the value/weight ratio of goods traded may be the fact that the technology content of the goods increases. Technology intensive goods, or fashion goods, tend to have a highly volatile market demand due to either short (technical) product cycles or fashion cycles. An increase of the share of technology or fashion goods in the total of the goods traded implies an increase in the demand for timeliness of transport services and for the speed of transport. As an example, Evan and Harrigan (2003) demonstrate that clothing apparel is an example where ideal characteristics are difficult to discern and therefore firms produce and ship close to sales dates. The requirement of short shipping periods had even led to the relocation of outsourced production to less distant countries. As Hummels has argued, the structure of traded goods changes towards more complex manufactured goods (Hummels, 2001a). For transport policy this has the consequence that the geography of freight flows changes, as do the associated demands for investment in transport infrastructure. The structural shift towards more time-sensitive goods is expected to be paralleled by a shift from ocean shipping and land transport towards air shipping, as it is associated with an increase of the value/weight ratio. Harrigan and Venables (2004) argue that time costs are qualitatively different from monetary trade costs in that they promote clustering of economic activities.
4. TRANSPORT COSTS, INTERNATIONAL TRADE AND ECONOMIC DEVELOPMENT
A second major discussion block during the Round Table addressed the question of how reduced transport costs lead to repercussions outside the transport sector. If these effects are important to the overall economy, it is vital that the planning and evaluation of transport policy measures takes account of these effects. This holds in particular for decisions to invest in transport infrastructure. As part of the benefits of transport policy measures that support international economic relations will not accrue to the domestic population but to firms and households in foreign countries, a lack of co-ordination between national and regional governments is likely to lead to a persistent under-investment in international transport facilities. The first part of the Round Table discussion led to the assessment that the costs of international transport had much less declined than was often presupposed. The second discussion block aimed at identifying the mechanisms by which the reduction of international transport costs benefits the overall economy. The dimension of the impact of international transport costs on national incomes depends first of all on the structure of the markets of traded goods. For perfectly competitive markets, the point of departure for an assessment of the income effects of reduced international transport costs is an economy whose exports are produced using imported inputs. High transport costs squeeze the domestic value added of such an economy in two ways: the domestic producer receives the world market price of the export good minus the (high) shipping costs for the export good, and pays a relatively full import price which includes the shipping costs for imported inputs. The numerical example given in the background paper of Venables shows how small changes in trade costs may have strong effects on domestic incomes. Given the typical capital and labour endowments of countries TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
SUMMARY OF DISCUSSIONS -
97
which have, at the same time, a poor transport infrastructure, the consequences of higher trade costs will have to be borne almost entirely by wage earners. How strong these effects are depends on the transport intensity of the economy (Venables and Limao, 2002). If peripheral countries typically export goods which have a low transport intensity, the wage squeeze resulting from high trade costs could be relatively unimportant. In fact, there seems to be a close correlation between how much skills are used in producing export goods and their transport intensity. For example, one of the fastest growing areas of world trade has been in parts and components, as firms outsource various stages of the production process. While individual components may be cheap to transport, this sort of activity is highly dependent both on imported inputs and on exporting the outputs. This is why the overall production is “transport intensive”. The destination countries of outsourcing may therefore be strongly affected by the wage squeeze resulting from trade costs. High international transport costs have stronger effects on national incomes the smaller the non-tradable sector and the larger the absolute disadvantages in the production of goods which are imported rather than produced domestically at low levels of trade costs. When (potential) export industries produce under increasing returns to scale in sectors with horizontally differentiated products, firms face the choice of where to locate a plant of minimum efficient scale. The level of transport costs then strongly affects location choices as well. The Round Table discussed how the additional dimension of firms’ mobility modifies the link between international transport costs and national incomes. In the framework of the theoretical models of economic geography, the effects of transport costs on the location decisions of firms are greatest when they are neither very high (location decisions are then dominated by local demand as trade volumes become very small) nor very low (firms are then indifferent as to where to locate). The better the market access of an economy, the higher will be nominal and real wages. Smaller countries are disadvantaged at intermediate levels of transport costs, being exposed to strong import competition, and will therefore benefit most from a reduction of trade costs beyond an intermediate level. High trade costs and low levels of international trade may have other negative consequences for national incomes and the catching up of poorer economies: with international trade, countries with relatively small R&D sectors may benefit from spillovers of technical and organisational knowledge (Coe and Helpman, 1995). Lower transport costs might also help Foreign Direct Investments if they are of a vertical type. Vertical FDI results from parts of domestic production processes being relocated to foreign countries with different relative factor prices. As is well known, foreign direct investment can also strongly support the international diffusion of technical and organisational knowledge (Barba Navaretti et al., 2004). The empirical evidence on the above relationship which was discussed by the Round Table suggests that there is a strong positive relationship between income and market access. For example, focussing on Europe, there is evidence of a wage gradient from Belgium/ Luxembourg (with the best foreign market access), through France and Britain to Spain, Portugal and Greece. Hypothetical experiments on the basis of the estimated relationship between market access and income show that being landlocked and being an island both have negative effects on income. Changing the countries’ trade openness from the 1994 value to the most open status possible led to extremely large income gains, of around 25 per cent, for countries that were relatively closed in the initial situation. To obtain a quantitative picture of the extent to which transport policy contributes to openness, the impact of infrastructure investments on openness was analysed. The infrastructure measure used was designed to gauge the costs of travel in and through a country. It was constructed as an average of TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
98 – SUMMARY OF DISCUSSIONS the density of the road network, the paved road network, the rail network and the number of telephone main lines per person. The results point to the importance of infrastructure in determining both trade costs and trade volumes. To quantify the magnitude of the infrastructure effects, the estimated elasticities were used to establish the costs and benefits of being at different points in the country distribution of infrastructure investment endowments. Improving the infrastructure from the median to the top 25th percentile reduces the cif/fob ratio from its median value of 1.28 to 1.12. This reduction could increase the trade volume by a full 68 per cent. The results of the theoretical and empirical analysis of the effects of trade costs on incomes and of transport policies, in particular infrastructure investment for trade costs, suggests that these effects deserve more attention in cost-benefit and policy analyses than they have received in the past. The analysis also shows that the costs of a failure to internationally co-ordinate transport policies, and in particular transport infrastructure investment, might have high opportunity costs in terms of real income and wages. To proceed from these conclusions to the implementation of trade cost-reducing measures requires an understanding of which measures are the most cost effective in which economic circumstances. It must also take account of the high resource requirements that some of the trade cost-reducing measures are associated with. The background paper provided by Wilson, Luo and Broadman, Trade and transport facilitation: European accession and capacity building priorities, aims at the identification of measures that belong to national policies which have a high potential for promoting international integration, in particular of the Western, Middle and Eastern European economies.
5. TRADE FACILITATION
Traditional definitions of “trade facilitation” have focussed on narrow measures to achieve the reduction of international trade costs. With the policy barriers to international trade being much reduced, further progress towards greater economic integration depends on improved efficiency in logistics at ports and customs but also streamlined regulatory policies, deeper harmonisation of standards and conformance to international norms so that overall transaction costs are lowered. Domestic reforms are at the centre of the current policy proposals for trade facilitation. They include greater transparency for customs procedures, ensuring that operational decisions are rules-based (rather than discretionary), enhancing the professionalism of customs officials, harmonising product and technical standards with international or regional regulations and strengthening the integration of new technologies into the transport and communications infrastructure. This holds as well and in particular for the transition countries of Europe, the CIS and Central Asia. The increased importance of the domestic reform steps to facilitate trade has been reflected in changes on the agenda of international trade negotiations. Trade facilitation was added to the policy dialogue on trade issues at the Singapore Ministerial of the World Trade Organization (WTO) in 1996. Moreover, in August 2004 the WTO decided to focus part of the negotiations currently underway in the Doha Round to trade facilitation issues. They are today at the centre of the Doha Development Agenda. As part of such negotiations, the European Union has been a leading advocate of discussing
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
SUMMARY OF DISCUSSIONS -
99
regulatory reform, modernisation of customs regimes and infrastructure investment related to lowering trade logistics costs. The empirical study which served as a background paper to the Round Table discussion was based on a paper by Wilson, Mann and Otsuki (2004). It uses four separate indicators to estimate trade gains due to trade facilitation progress in: – – – –
port efficiency; customs regimes; regulatory policy; and information technology infrastructure.
On a global basis, the study of Wilson, Mann and Otsuki (2004) suggests that increased port efficiency, improved customs regimes and streamlined and/or harmonised regulatory policies can lead to substantial trade gains. Their analysis indicates that, for the 75 sample countries in the global study, raising capacity halfway to the global average would yield a US$377 billion gain to world trade. In the background paper of the Round Table, the same analytical approach was applied to eight new members of the EU, and four accession candidate members, Bulgaria, Croatia, Romania and Turkey. The analysis uses a gravity model of bilateral trade flows. The estimated gravity equations are used to simulate prospective country-specific measures to facilitate trade in the four dimensions listed above. Indicators are constructed to measure the degree of trade facilitation in the sample countries relative to benchmark countries, which have been defined using data from the World Competitiveness Yearbook (IMD) in 2000. Singapore was the best performer in port efficiency and Finland the best performer in the other three areas. For each country, the indicator “port efficiency” is an average of the efficiency of port facilities, inland waterways facilities and air transport. The indicator “customs regimes” captures the hidden import barriers other than published tariffs and quotas, and the irregular extra payments or bribes connected with import and export permits. The indicator “regulatory” policy is constructed as the average of the transparency of government policy and the control of corruption. Finally, the indicator “information technology infrastructure” is a measure of the speed and cost of Internet access and the contribution of the Internet to the reduction of inventory costs. Inevitably, the “operationalisation” of policy variables, for areas as broad and diverse as those to measure trade facilitation, invites critical discussion. Potential improvements of the indicators notwithstanding, the results of the empirical analysis are highly illustrative of the achievements and remaining actions required by the countries in the sample. The analysis revealed that EU-8 member countries still have deficits with respect to the four areas of trade facilitation. As to three candidate member countries (Bulgaria, Romania and Turkey), the development of their trade facilitation is further behind, the state of their customs regimes being 58 per cent of the level of the EU-15 countries. There are substantial differences between the EU-8 countries; Estonia being a good performer with respect to port efficiency, IT infrastructure and customs efficiency. Major differences also exist between the candidate accession countries with respect to the variables indicating the state of trade facilitation. Controlling for development differences between the countries, the analysis of the data suggests that the new and candidate EU member countries are, in general, weak performers in all four of the trade facilitation dimensions examined. Even in the context of their relatively low level of development, their trade facilitation development is under the benchmark level. The only exception is Estonia, which performs more strongly than the benchmark level in all four categories.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
100 – SUMMARY OF DISCUSSIONS By making use of the estimated gravity equations, the role of trade facilitation in bilateral trade is examined. If infrastructure is upgraded and transaction costs lowered, trade volumes can expand. Expanded trade could then contribute to higher incomes, as the Round Table had debated in the second discussion block, to higher growth rates and to a lower level of unemployment. Moreover, the implementation of trade facilitation measures could increase the reliability of delivery times. The empirical gravity equations are used to simulate the impact of the hypothetical improvements in port efficiency, customs regimes, regulatory policy and information technology infrastructure to half-way the benchmark level on bilateral trade flows. After controlling the effects of tariffs, differences in development level, distance between trading countries and the regional characteristics of exporters and importers, the analysis shows that behind-the-border factors do play a critical role in determining bilateral trade flows. Three out of the four policy categories (port efficiency, regulatory regimes and IT infrastructures) increase exports relatively more than imports. The new and candidate member countries of the EU would expect large trade gains as well as an improvement in the balance-of-payments situation. The greatest absolute trade gains, of US$49 billion and US$62 billion, would result if port efficiency and IT infrastructure reached half the average level of the EU-15. Seventy per cent of these gains would result from export expansion. Trade gains would result in the new member and candidate member countries, as well as the EU-15 countries, benefiting from the unilateral actions of the EU-8 and candidate accession countries. Among the four dimensions of trade facilitation, improvement of IT infrastructure will result in the highest trade gains (more than US$4 billion), which is greater than the gains from port efficiency (with trade gains close to US$3 billion). Improvements in regulatory policy and customs regimes lead to about the same trade gains, of around US$1.5 billion each. That is to say, more than 40 per cent of the trade gains come from improvements in IT infrastructure and almost 30 per cent from port efficiency.
NOTE
1.
As Hummels (2001) has reported, a high share of the observations is imputed, severely limiting the value of the data for analytical purposes.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
SUMMARY OF DISCUSSIONS -
101
BIBLIOGRAPHY
Anderson, J.E. and E. v. Wincoop (2004), Trade costs, Journal of Economic Literature, 42: 691-751. Barba Navaretti, G., A. Venables and F.G. Barry (2004), Multinational Firms in the World Economy, Princeton. Brun, J.-F., C. Carrere, P. Guillaumont and J. de Melo (2002), Has distance died?, CEPR Discussion Paper 3500, London. Coe, D. and E. Helpman (1995), International R&D spillovers, European Economic Review, 39: 859-887. Coe, D.T., A. Subramanian and N. Tamirisa (2002), The missing globalisation puzzle, IMF Working Paper, Washington. Evans, C.L. and J.E. Harrigan (2003), Distance, time and specialization, NBER Working Paper 9729, Cambridge, Mass. Feenstra, R.C. (1998), Integration of trade and disintegration of production in the global economy, Journal of Economic Perspectives, 12: 31-50. Gallup, J. and J.D. Sachs (1999), Geography and economic development, in: B. Pleskovic and J. Stiglitz (eds.), Annual World Bank Conference on Development Economics, Washington, DC. Grossman, G. (1998), Comment on Deardorff, in: J.A. Frankel (ed.), The Regionalization of the World Economy, Chicago. Harrigan, J. and A.J. Venables (2004), Timeliness and agglomeration, CEPR Discussion Paper 4249, London. Hummels, D. (2001), Time as a trade barrier, mimeo, Purdue University, Lafayette. Hummels, D. (2001), Toward a geography of trade cost, mimeo, Purdue University, Lafayette. Hummels, D. and A. Skiba (2004), Shipping the good apples out? An empirical confirmation of the Alchian-Allen conjecture, Journal of Political Economy, 112: 1384-1402. IMD, World Competitiveness Yearbook, Lausanne. Obstfeld, M. and K.S. Rogoff (2000), The six major puzzles in macroeconomics: Is there a common cause?, NBER Macroeconomics Annual, 15: 339-390.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
102 – SUMMARY OF DISCUSSIONS Portes, R. and H. Rey (2002), The determinants of cross-border equity transaction flows, London Business School Working Paper, London. Radelet, S. and J.D. Sachs (1999), Shipping Costs, Manufactured Exports and Economic Growth, Harvard Institute for International Development, Cambridge, Mass. Venables, A. and N. Limão (2002), Geographical disadvantage: a Heckscher-Ohlin-von Thünen model of international specialisation, Journal of International Economics, 58: 239-263. Wilson, J.S., C.L. Mann and T. Otsuki (2004), Assessing the potential benefit of trade facilitation: A global perspective, World Bank Working Paper 3224, Washington, DC.
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
SUMMARY OF DISCUSSIONS -
103
LIST OF PARTICIPANTS
Prof. Roger VICKERMAN University of Kent Dept. of Economics Keynes College GB-CANTERBURY, CT2 7NP Royaume-Uni
Chairman
Mr. Harry BROADMAN Senior Economist World Bank 1818 H Street, NW WASHINGTON DC 20433 États-Unis
Co-rapporteur
Mr. John S. WILSON World Bank 1818 H Street, NW WASHINGTON DC 20433 États-Unis
Co-Rapporteur
Assoc. Prof. David HUMMELS Associate Professor of Economics Purdue University Krannert School of Management Krannert Building 403 West State Street WEST LAFAYETTE, IN 47907-2056 États-Unis
Rapporteur
Prof. Anthony VENABLES London School of Economics and Political Science (LSE) Houghton Street GB-LONDON WC2A 2AE Royaume-Uni
Rapporteur
Mr. Faik S. ALAKBAROV Chief, Transport Policy & Economics Dept. Ministry of Transport Block 1054 Tbilisi avenue AZ-370602 BAKU Azerbaïdjan TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
104 – SUMMARY OF DISCUSSIONS Prof. Michel BEUTHE Groupe Transport et Mobilité (GTM) Facultés Universitaires Catholiques de Mons (FUCAM) 151 Chemin de Binche B-7000 MONS Belgique Prof. Alain BONNAFOUS Directeur de l'ISH Laboratoire d'Économie des Transports (LET) ISH 14 avenue Berthelot F-69363 LYON Cedex 07 France Professor Michael BROWNE University of Westminster Transport Studies Group 35 Marylebone Road GB-LONDON, NW1 5LS Royaume-Uni Prof. Kenneth BUTTON George Mason University School of Public Policy FAIRFAX, VA 22030 États-Unis Ms. Nadia CAID Administrator Sectoral Policy Integration National Policies Division Environment Directorate OECD Prof. Yücel CANDEMIR Director, Technological and Economic Dev. Research Centre Istanbul Technical University Istanbul Teknik Üniversitesi (ITU) TR-80680 Macka/ISTANBUL Turquie Dr. Pierre Philippe COMBES GREQAM et CERAS - ENPC 2, rue de la Vieille Charité F-13002 MARSEILLE France
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
SUMMARY OF DISCUSSIONS -
Mr. Elmar FARAJOV Head of TRACECA Program Unit Ministry of Transport International Relations Department 1054 Tbilisi Prospekt AZ-1122 BAKU Azerbaïdjan Prof. Kazimierz FIEDOROWICZ Adviser to the Minister Ministry of Infrastructure Chalubinskiego St. 6 PL-00-928 WARSAW Pologne Prof. Marc GAUDRY (Université de Montréal) INRETS 2 av. du Général Malleret Joinville F-94114 ARCUEIL Cédex France Mr. Bernard HOEKMAN Research Manager International Trade Section Development Research Group World Bank 1818 H St. N.W. WASHINGTON DC 20433 États-Unis Mr. Jan HOFFMANN Economic Affairs Officer Substantive Division UNCTAD Palais des Nations CH-1211 GENEVA 10 Suisse Mr. Anthony KLEITZ Head of Division Trade Liberalisation and Review Trade Directorate OECD
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
105
106 – SUMMARY OF DISCUSSIONS Dr. Atsushi KOIKE Visiting Researcher TNO INRO Schoemakerstraat, 97 P.O. Box 6041 NL-2600 JA DELFT Pays-Bas Mme Evdokia MOISE-LEEMAN Administrator (Regulatory Issues, Trade Facilitation) Trade Liberalisation and Review Trade Directorate OECD Dr. Hildegunn Kyvik NORDÅS Institute for Research in Economics and Business Administration (SNF) Breiviksveien 40 N-5045 BERGEN Norvège Prof. Elisenda PALUZIE Facultat d'Econòmiques Universitat de Barcelona Av Diagonal E-690 08034 BARCELONA Espagne Mr Attila RAJMON Institute for Transport Sciences (KTI) Than Karoly ut. 3-5 PO Box 107 H-1518 BUDAPEST Hongrie Prof. Werner ROTHENGATTER Universität Karlsruhe Institut für Wirtschaftspolitik und Wirtschaftsforschung (IWW) Postfach 69 80 D-76128 KARLSRUHE Allemagne Prof.Dr. Wlodzimierz RYDZKOWSKI Chairman of Department University of Gdansk Department of Transportation Policy Armii Krajowej 119/121 PL-81-824 SOPOT Pologne
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
SUMMARY OF DISCUSSIONS -
Prof.Dr. Karl STEININGER University of Graz Department of Economics Universitätsstrasse 15 A-8010 GRAZ Autriche Prof. Siri Pettersen STRANDENES Norwegian School of Economics and Business Administration Centre for International Economics and Shipping (SIOS) Helleveien, 30 N-5035 BERGEN-SANDVIKEN Norvège Dr. L.A. TAVASSZY TNO INRO Schoemakerstraat, 97 P.O. Box 6041 NL-2600 JA DELFT Pays-Bas Prof. Eddy VAN DE VOORDE Universitaire faculteiten St Ignatius Te Antwerpen Department of Transport and Regional Economics Prinsstraat 13 B-2000 ANTWERP 1 Belgique
ECMT SECRETARIAT Mr. Jack SHORT, Secretary General JOINT OECD/ECMT TRANSPORT RESEARCH CENTRE Dr. Andreas KOPP Chief Economist Dr. Michel VIOLLAND Administrator Mrs. Julie PAILLIEZ Assistant Ms Françoise ROULLET Assistant
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
107
108 – SUMMARY OF DISCUSSIONS
ALSO AVAILABLE
Managing Commuters' Behaviour - A New Role for Companies. Series ECMT – Round Table 121 (2002) (75 2002 11 1 P) ISBN 92-821-1299-3 Transport and Exceptional Public Events. Series ECMT – Round Table 122 (2003) (75 2003 04 1 P) ISBN 92-821-1305-1 Vandalism, Terrorism and Security in Urban Passenger Transport. Series ECMT – Round Table 123 (2003) (75 2003 07 1 P) ISBN 92-821-0301-3 Transport and Spatial Policies: The Role of Regulatory and Fiscal Incentives. Series ECMT – Round Table 124 (2004) (75 2004 09 1 P) ISBN 92-821-2321-9 European Integration of Rail Freight Transport. Series ECMT – Round Table 125 (2004) (75 2004 06 1 P) ISBN 92-821-1319-1 Airports as Multimodal Interchange Nodes. Series ECMT – Round Table 126 (2005) (75 2005 03 1 P) ISBN 92-821-0339-0 16th International Symposium on Theory and Practice in Transport Economics. 50 Years of Transport Research (2005) (75 2005 05 1 P) ISBN 92-821-2333-2 Time and Transport. Series ECMT – Round Table 127 (2005) (75 2006 01 1 P) ISBN 92-821-2330-8 National Systems of Transport Infrastructure Planning. Series ECMT – Round Table 128 (2005) (75 2005 10 1 P) ISBN 92-821-2341-3 Transport Services: The Limits of (De)Regulation. Series ECMT – Round Table 129 (2006) (75 2006 02 1 P) ISBN 92-821-2345-6 To register for information by email about new OECD publications: www.oecd.org/OECDdirect For orders on line: www.oecd.org/bookshop For further information about ECMT: www.cemt.org
TRANSPORT AND INTERNATIONAL TRADE – ISBN 92-821-1338-8 - © ECMT, 2006
OECD PUBLICATIONS, 2, rue André-Pascal, 75775 PARIS CEDEX 16 PRINTED IN FRANCE (75 2006 13 1 P) ISBN 92-821-1338-8 – No. 55229 2006
Page 1
13 0 ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT
TRANSPORT AND INTERNATIONAL TRADE With the removal of many trade policy barriers, further international economic integration depends largely on the reduction of trade costs originating in the transport sector.
Background papers were provided by David Hummels (Purdue University), Anthony Venables (London School of Economics and Centre for Economic Policy Research) as well as Harry Broadman and John S. Wilson (World Bank).
TRANSPORT AND INTERNATIONAL TRADE
The Round Table discussion focused on the structure and development of international transport costs over the past decades and the benefits to be expected from investment in international transport facilities and the reduction of the costs of crossing borders.
130
R O U N D TA B L E
(75 2006 13 1 P1) ISBN 92-821-1338-8
-:HSTCSC=VVXX]\:
2006
C E N T R E
16:26
R E S E A R C H
27/07/06
T R A N S P O R T
fullTab130GB
EUROPEAN CONFERENCE OF MINISTERS OF TRANSPORT
TRANSPORT AND INTERNATIONAL TRADE
ROUND TABLE
13 0