Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
The significance of freight transport
Although the transport sector as a whole is receiving attention in the environment-development and eco-restructuring discussion - in particular, in respect to the future of private passenger transport the discussion has tended to focus on the central role of the car and on urban settings. It is true that Western industrial societies and their cities can be characterized as autocentric. It is also true that in terms of perhaps the most topical environmental policy concern on the international agenda - climatic change - the contribution of the automobile to carbon emissions and the build-up of atmospheric concentrations of greenhouse gases is greater than that of freight transport. None the less, the relative importance of the two - cars and trucks - must be considered in terms of the significance of the sector as a whole in commercial energy use and its overall contribution to anthropogenic carbon emissions.
Transport energy use accounts for more than one-quarter of worldwide commercial energy consumption (Grübber 1993). Moreover, with an average growth rate of 2.7 per cent per annum between 1971 and 1990, transportation energy demand is the fastest-growing energy end-use category (IEA 1993a,b). Passenger travel accounts for 60 per cent of transport energy use, and freight movement for the remaining 40 per cent (Nakicenovic 1993). Transport energy use accounts for almost 30 per cent of all energy end-use-related anthropogenic carbon dioxide (CO2) emissions (OECD 1995). Since anthropogenic CO2 emissions currently amount to 26 billion tonnes annually, of which 80 per cent are energy related, transport is responsible for around 7.5 billion tonnes of CO2 per annum. Passenger cars account for just less than half of this (around 3.6 billion tonnes) and goods vehicles for about one third (c. 2.4 billion tonnes). Other vehicles account for the rest.
Although the freight share of CO2 emissions is less than the passenger car share, the freight share is increasing. Moreover, at least in respect to most OECD countries, the energy intensity of goods transport is increasing owing to shifts to faster modes (Schipper and Meyers 1992). The Second Assessment Report of the Intergovernmental Panel on Climate Change (IPCC) has discussed this development in some detail and has analysed the reasons for it in respect to geographical differences between countries and to the types of trucks involved. "In countries where services and light industry are growing faster than heavy industry, the share of small trucks and vans in road freight is increasing. The energy intensity of light trucks is high compared with large trucks. Along with general increase in the power-to-weight ratios of goods vehicles, these developments offset and in some cases outweigh the benefits of improving engine and vehicle technology" (IPCC 1996, pp. 690-691). The two noteworthy points are that transportation energy intensity tends to be lower in countries where a high proportion of goods traffic is made up of bulk materials or primary commodities than in more industrialized and diversified economies, and that improvements in vehicle technology have partly been used to increase vehicle performance rather than to reduce energy use. Both contribute to the increasing energy intensity of goods transportation within the OECD countries.
Freight transport is therefore absolutely significant in its own right as a source of pollution and a target for eco-restructuring attention. Perhaps most important is to compare freight CO2 emissions with those of industry. Industrial activity as a whole accounts for just less than 35 per cent of all anthropogenic carbon emissions; i.e. around 9 billion tonnes of CO2 per annum. The contribution by freight transport is therefore equivalent to more than a quarter of all industrial CO2 emissions. Moreover, the freight contribution is greater than that of any single industrial sector and is increasing. It is responsible for 10 times the energy-related CO2 emissions of the cement industry (although only just over twice that industry's overall CO2 emissions), eight times those of the pulp and paper industry, and twice those of the steel industry.
As well as contributing to carbon emissions, freight transport is associated with other pollutants. These include nitrogen dioxide (NO2) and sulphur dioxide (SO2). NO2, like CO2, is a greenhouse gas - though more than 200 times more potent than CO2 in terms of radiative forcing. Both NO2 and SO2 contribute to environmental acidification. Carbon monoxide (CO) and nitrogen oxides (NOx), both of which are also emitted, have reaction products that add to these effects. The quantities of these emissions have recently been estimated as 2.1 g of CO and 1.85 g of NOx per tonne-kilometre (ECMT 1995). Noise pollution and accident fatalities were also included in the ECMT (European Conference of Ministers of Transport) study and are important because, although not directly relevant to globally or regionally threatening environmental change, they, too, represent costs that remain to be fully internalized within fuel, road, and transport prices. Another cost of transportation is the time cost of delays caused by congestion. Internalization of the costs associated with these externalities would contribute to reducing freight transport volumes and so, also, to eco-restructuring.
Viewed against the backdrop of the role that freight transport currently plays in environmental change, its relative neglect within the eco-restructuring discussion is surprising. Transport is usually considered in terms of passenger transport and car use, in respect to which discussions have focused heavily on technological opportunities for reducing unit impacts. These have centred around the need for energy efficiency improvements and for low-carbon or carbon-free energy carriers as replacements for fossil fuels. Work has also looked at public alternatives to private transport (train, metro, bus, tram, light railway, etc.) and at how modal shifts might be encouraged. In terms of demand management, there has been some discussion of the technological opportunities to substitute for passenger travel; for example, by using telecommunications for remote working, teleshopping, and home entertainment.
Relatively little has been done, however, to examine the relationships between patterns of space use and passenger transport demand; for example, the links between a greater mixing of land uses (residential, recreational, retail, office, etc.) or between the scale of provision of services (large centralized schools, hospitals, and shops versus small-scale facilities serving local neighborhoods) and auto use. This omission is significant in respect to the parallels that exist between passenger transport (which is widely studied in the debate) and freight transport (which is not), because considerable potential for demand reduction in the freight sector could lie in reorganizing spatial divisions of labour and logistics to reduce transport intensities at the product level.
Several important issues in respect to past and future trends in the development of freight transport are, thus, raised:
- What factors and trends have been important in the past patterns of freight transport growth and development?
- What is the current level and structure of production of freight transport?
- What factors and trends are shaping future developments in freight transport volumes and patterns?
- How much scope is there for reducing freight volumes?
- How might this scope be taken up?
- How would the space-use, logistics, transport, and trade patterns of a sustainable society and economy differ from today's patterns?
The remainder of this chapter is devoted to these questions and is structured around them. The discussion centres mostly on industrialized countries because these face the task of shifting from deeply embedded and unsustainable land-use patterns and reducing their transport dependence. Many of the examples and data given reflect European Union (KU) experience. None the less, these are believed to be representative for other industrialized countries. Inferences can be drawn for developing countries, especially in respect to opportunities for avoiding heavy investments in infrastructures that imply long-term commitments of environmental (and also economic) resources to potentially unproductive or low-productivity uses.
Past growth and patterns of freight transport development
Present-day patterns of transport, logistics, and land use are the still evolving outcome of a century-long - and still ongoing - process of industrial development. Consistent features of the trajectory have been its exploitation of:
- volume production and consumption as a metric and target for economic growth,
- cost reduction via productivity improvements as an engine for growth,
- scale economies and comparative advantage for total factor productivity gains,
- energy/materials/capital to substitute for labour (for labour productivity gains).
Development has followed an energy- and materials-intensive production regime based heavily upon the one-time conversion of natural (environmental) capital stock to current income. The generated income has been used to develop different forms of "man-made" capital appropriate to the continuation of the trajectory and to provide for high levels of current consumption.2
Within this trajectory, the most important factors in shaping the evolving geography of economic activities have been: deregulation and the liberalization of markets; improvements in transport infrastructures/technology; the maintenance of low transport and energy prices; and developments in logistics and in information technologies. Competition and the search for profits have been the major dynamic forces. The major constraints have been some (usually weak) restrictions on land development and land use.
Market liberalization and deregulation
Liberalization has taken place within countries, between pairs or groups of countries, and at the global level, affecting both tariff and non-tariff barriers in respect to markets in factors, goods, and services. All factors (save labour) have been affected. The freeing-up of capital markets has been particularly comprehensive. The process began immediately after the end of World War II in an attempt to stimulate economic growth and prosperity at the world level and avoid a repeat of the depression that had accompanied inter-war protectionism. Some changes were agreed bilaterally among major trading partners, but the major multilateral mechanism for liberalizing world trade has been the General Agreement on Tariffs and Trade (GATT) process, which started in 1947 and culminated in the Uruguay Round of negotiations and Marrakesh agreements. The development of regional trading blocs has also been a significant - in the case of West European countries, perhaps the most significant development in the post-war liberalization process. Blocs now include the European Union, the North American Free Trade Area, the Association of South-East Asian Nations, and Mercosul.
By way of illustrating the removal of tariff barriers, there is no better example than the passenger car. The contrast between tariff levels during the highly protectionist period of the 1920s and 1930s and those of the 1970s and 1980s is stark. During the inter-war period each European country separately determined its tariff level. The average across the major European carproducing nations in the 1930s was around 45 per cent of customs value. In the case of France and Italy, tariffs were in the range 47-74 per cent and 101-111 per cent, respectively. Japan also had a highly protected market. Its tariff level in 1940 was 70 per cent. By the early 1980s, the situation was completely different. The EU countries had harmonized on a tariff structure within a range to 10.5 per cent of customs value. Tariffs in the United States and Japan were minimal to non-existent (Altschuler et al. 1984).
The impact of deregulation and trade liberalization on transport demand arises through three principal effects: the exploitation of comparative advantage; the exploitation of scale economies; and the expansion of markets. In respect to the first two, the removal of barriers increases geographical specialization and concentration of production. Comparative advantage is particularly important for inter-industry trade, whereas exploitation of scale economies or experience is particularly important for intra-industry trade. In respect to the third, falling import barriers, lower product prices, higher per capita consumption, and access to wider markets lead to increases in the average distance between points of production and points of consumption (Gabel 1994).
As barriers to factor mobility have been reduced or removed, the world economy has become steadily internationalized. Together with accompanying technological, institutional, and organizational innovations, this has allowed economic activities to be restructured through a series of "spatial fixes" (Harvey 1982). As the spatial framework has expanded, entrepreneurs have been free to engineer the most profitable business arrangements under the prevailing new conditions. Harvey argues that, through these, a long-threatened growth crisis has been repeatedly postponed. Each new spatial fix has brought higher factor productivities and has permitted growth to continue within the context of the same mass production/mass-consumption model.
Transport improvement, supply, and cost
Governments have taken responsibility for infrastructure provision and have used infrastructure development in macroeconomic policy, employment creation, regional development, and defence strategy as well as in transport policy. Governments have typically taken a supply expansion rather than a demand-side management approach to balancing demand and supply. More important, they have catered for "anticipated" demands. A result has been the considerable extension and improvement of road infrastructures. Within OECD-Europe, for example, the motorway network length has almost tripled since 1970. Within the OECD group as a whole, motorway network length in 1993 stood at 150,266 km. This represents a 34 per cent increase since 1980. The average motorway network density within the OECD is currently 0.44 km/100 km2; within some European countries (Belgium and the Netherlands) it is more than 5.5 km/100 km2. Coupled with progress in truck technology, motorway development has significantly improved average speeds and journey times. The truck fleet has grown by more than 50 per cent since 1980. In 1993, there were more than 98 million goods vehicles in service within OECD countries. Within some countries of the EU (e.g. France, Italy, and the Netherlands), the fleet size has doubled since 1980.
Transport and energy prices
Meanwhile, markets in respect to energy, land, and transport have been so constructed as to externalize many costs. The environmental costs of energy use in terms of depletion and pollution have not been factored into energy prices, which have in any case been low and either stagnating or reducing in real terms over the past decade. Diesel and petrol prices to consumers were lower in 1993 than in 1985 in several countries (e.g. the United States). This reflects depressed oil prices, which have only partly been offset by increases in the percentage share of taxes in consumer prices. Neither are the environmental and social costs of land-take factored into user charges for roads, ports, or airports. As with all public subsidies, the benefits are available to be privatized while the costs are socialized. This has provided strong incentives for businesses to increase the use of transport and to make logistics a key aspect of technological choice in respect to strategies for rationalization, marketing, and competitiveness. Part of the low cost of transport has been passed on to consumers in the form of transport-intensive goods and services with prices lower than would have been possible without the subsidies.
Positive feedback mean that many of these distortions are self reinforcing. Subsidy has favoured high levels of mobility and low density land-use patterns. High dependence on mobility has been important, in turn, as a justification for further infrastructure provision, particularly because of the inherent biases toward provision within the benefit-cost and discounting methodologies used by government agencies for investment appraisal.
Developments in logistics
Particularly since the 1980s (although the start of the development can be traced back to the 1970s) the function of logistics within business operations and competitive strategy has changed. Key drivers have been shifts in consumer markets, with less brand loyalty and greater competition among producers. This has elevated the importance of customer service - including product availability, delivery lead-times, reliability of deliveries, etc. - within the competitive process and led to shifts in approaches to business management and logistics. A related development has seen logistics reorganization being used to secure reductions in inventory and warehousing costs. Facilitating developments have been deregulation and privatization and rapid advances in information technologies.
Change has followed a three-stage development (Cooper 1995). In the 1970s, companies began to bring together transport, warehousing, and inventory management at the firm level (integrated distribution). The concept was then extended beyond the boundaries of the firm so that the whole materials chain from the sourcing of raw materials to the final consumption of finished products was integrated (channel integration). By the early 1990s, a third phase has seen operations becoming integrated across national boundaries (geographical integration), which, as well as enabling the promotion of new product strategies, offers companies the opportunity to reap scale economies in purchasing, manufacturing, warehousing, and transport. By regionalizing warehouses, firms are able to rationalize and reduce inventory. A parallel development upstream in the supply chain has been a growing emphasis on "focused factories," where companies select factories to specialize in producing parts of the product range rather than its entirety. Focused factories invariably serve more than one country.
Information technology
An equally important development, from the perspective of truck traffic generation, has been the growing importance of information technology in logistics management. Whereas inventory holding levels used to be greater than strictly necessary to compensate for deficiencies of information, transparency in the supply chain by virtue of information technology means that it is increasingly possible to reduce levels of inventory (Cooper 1995). Accurate and timely information on stock holdings and whereabouts increasingly encompasses stock on trucks. Together with a more responsive transport service, remaining inventory can be concentrated at a smaller number of strategic locations. Firms have been able profitably to trade off higher transport operating costs against (much) lower warehousing costs.
Within Europe, the lead companies in the integration process are those in the automotive and business equipment sectors together with some retailers (e.g. IKEA and Benetton). Bosch-Siemens, for example, has recently consolidated its Scandinavian warehousing operations - formerly based upon distribution centres in Finland, Norway, and Sweden - into one site in southern Sweden. Unilever (household products) and Ciba-Geigy (pharmaceuticals) are leaders in the focused factory approach. As well as implying higher transport demands and more international transport, these developments depend heavily upon road freight, which is much better suited to meet the new demands of users than competing modes. An important consequence has been the generation of higher levels of road freight activity through increases in average haulage length.
Spatial and transport outcomes
The most important spatial-economic outcomes of these developments and their implications for transport have been:
- increasing geographical specialization and concentration in production,
- growing separation of points in value-adding chains,
- increasing distance between points of production and points of consumption,
- internationalization of economic activities and of transport, growing transport volumes,
- increasing average haulage distances, shifts to faster modes.
In terms of the transport implications, the most significant effect has been rapid growth in the volume of truck-kilometres. In 1993, the OECD truck fleet travelled a total of 1,823 billion kilometres. This represents a 68 per cent increase over the 1980 kilometrage. Truck movements on important international corridor routes such as the Rhine and Danube valleys and through sensitive Alpine passes have become particularly problematic.
Part of the overall growth in truck-kilometres is attributable to increases in average haulage length. Within the KU, the average haulage length of goods (all commodities) transported by road grew by more than 20 per cent in the 10 years to 1990. Some of the growth has also come from mode shifts. Competition favours the use of flexible, direct, and fast modes of transport. Truck and air have gained relative to slower and less direct modes. Rail, in particular, has lost out. Not only is the rail share of total tonne-kilometrage reducing, but rail has experienced absolute reductions in freight-tonne kilometres. Studies of German freight transport have shown that the threshold of substitution between slower modes and truck occurs at value densities of little more than DM 1-2/kg. Truck dominates the value range from this to the DM 100/kg level (all figures in 1987 DM), which encompasses most consumer products including autos and domestic electrical (white) goods. Above DM 100/kg, air freight becomes dominant (Grübber 1988).
International trade has grown consistently and has outpaced the growth in real world product. No EU country now imports less than 30 per cent of its total GNP.
Competition and profit-maximizing behaviours have been the critical driving force behind these impacts. Together, these have constituted a powerful imperative to cut costs, expand market presence, and respond quickly to demand.
Evidence is rife of bizarre transportation consequences (Whitelegg 1994). Freight transport surveys have shown that, to take advantage of cost differences between countries, some products are shipped long distances in the process of adding value, ultimately only to arrive back at their point of origin. Prawns landed in Hamburg and destined to be sold in the German market have been found in transport to and from Poland, where they are cooked and peeled. Most West European household pot plants, wherever grown, are sold at auctions held in the Netherlands from where they are trucked to final market, sometimes even to places where they were originally grown. In both these cases, it is the same (not just an equivalent) product that is being transported in both directions. Even low value-density commodities and products have to be transported into foreign markets at risk of their producers losing overall market share. At the high value density end of the market, there are even more questionable, but apparently profitable and cost-effective, goods movements. An example is the case of the Cadillac Allante. The car body is manufactured in Italy. But the car is assembled and sold in the United States. The body is transported over 5,000 km by air from Turin to Detroit (Grübber and Nowotny 1990).
There is evidence that transport intensities at the national level (measured as tonne-kilometres per unit GNP) may now have peaked in some OECD countries and even have begun to decline. This is the result of higher average specific values (value/kg) of cargo and is likely to reflect changes in the structure of economic activity within countries - including the offloading of environmentally damaging aspects of production to other countries. It is unlikely, therefore, to be a reliable indicator of shifts to or away from transport dependence. Time-series of product-specific transport intensities would be better indicators but these are not routinely calculated. In an exceptional (and exemplary) study aimed at establishing transport intensity at the product level, Böge has used lifecycle methods to trace all the inputs used in the manufacture of a single pot of strawberry yoghurt emanating from a production plant in southern Germany (Böge 1993, 1995).
The study includes all inputs involved in the manufacture of the product and its packaging; e.g. strawberries, sugar, milk, glass containers, cardboard and paper labels, and aluminium container lids. The transport requirement associated with producing the inputs, bringing these to the production plant in southern Germany, and distributing the final product to consumers is summarized over the life cycle on a unit product basis. Böge expresses the transport intensity in terms that enable it to be considered - and possibly listed like any other "ingredient" on the label of the product. Each 150g strawberry yoghurt was found to have required the equivalent of more than 9 metres of truck movement. Every tonne of yoghurt sold accounts for more than 600 truck-kilometres. In respect to overall sales of yoghurt - a seemingly simple, healthy, and "natural" product marketed in massive quantities - the implied daily, weekly, or annual transport requirement is enormous. Production from the single plant studied, which serves mostly local markets, gives rise to 24,000 truck-kilometres annually.
Future developments affecting freight volumes and patterns
As concerns reducing the environmental impact of the freight transport sector, virtually all present trends point in entirely the wrong direction. Freight volumes (measured in tonne-kilometres) are increasing not decreasing. Moreover, the growth in freight volumes is now attributable almost entirely to increases in the distance that goods are transported - reflecting greater separation between points in the value-adding chain - rather than to increases in the quantity of goods transported. What is worse, transport volume per capita tends to increase as societies become more affluent and, even though transport intensity (measured per unit of GDP) is beginning to fall in the most affluent societies, the decrease is more than offset by increases in GDP per capita. What matters from an environmental standpoint is that freight transport volumes and energy intensities are increasing in absolute terms.
The trends at international level are also bad. Most developing countries assisted by Western advisers and consultants, by various international agencies, and by development banks - have development strategies that are heavily based upon providing physical infrastructures to facilitate industrialization on the Western model. If successful, these would lead to their joining the industrialized West in having economies and societies that are transport intensive and fossil fuel dependent. Shachar points out that this convergence of industrialized and developing countries on the same mobility patterns and fossil-energy-based economies is one of the main reasons - if not the main reason - for our moving away from sustainability at the global level (Shachar 1991).
There is also reason for concern over the effects of further liberalization of world markets as a result of the Marrakesh agreements that concluded the final, Uruguay Round of the General Agreement on Tariffs and Trade negotiations in April 1994. Although, in principle, trade liberalization is not inconsistent with environmental protection (because measures in respect to both have a common goal in securing efficient resource allocations), the sequencing of measures is important. Although it has been pointed out that both liberalization and environmental protection policies have broadly equal claims to priority - in that barriers to free trade and environmental externalities are both alleged to cause similar allocation distortions (Repetto 1993)3 - progress toward free trade is being made more rapidly than progress toward internalizing environmental externalities. Moreover, the nature of the two distortions is fundamentally different in that tariff and non-tariff restrictions have the effect of restricting the spatial range of economic operations, both production related and marketing. In so doing, they act to limit the distortions in resource allocations - especially the excessive use of transport - that arise from market imperfections and subsidies.
Gabel (1994) has analysed the likely effects of further liberalization on freight transport, addressing the impacts arising from several different linkages. Liberalization within the transport sector itself will increase competition and lower the prices charged by operators. Liberalization within other sectors - specifically agricultural commodities, chemicals, petroleum, coal, steel, metals, and automobiles presents considerable scope for trade expansion because these have traditionally been tightly protected at national level and are transport intensive. Liberalization of the transport equipment and energy industries will likely lead to reductions in the costs of inputs relevant to the supply of transport services and therefore to reductions in transport costs. Increased geographical specialization and concentration of production will increase average haulage lengths. Inter-industry trade is more likely to increase average transport distance than intra-industry trade because the source of trade gains - factor differences and comparative advantage - is geographically determined (Gabel 1994, p. 164). The impact of economic growth and increases in per capita income is complex. Of several, partially offsetting effects, that with the greatest environmental significance is likely to be the continuing displacement of bulk commodities by lower lot sizes with higher value-to-weight ratios. Although qualifying his analysis in several ways, Gabel concludes: "In summary, virtually all of the changes anticipated to accrue from trade liberalisation will probably increase transport output" and "may lead to a shift in the mode of transport to road and air" (Gabel 1994, pp. 169 and 170).
Results from studies aimed at projecting the impact of recent EU liberalization and deregulation show similar results (e.g. Venables and Smith 1988; Gabel and Roller 1993). Based on an econometric model, Gabel and Roller forecast impacts on international transport movements of the complete elimination of internal (non-tariff) trade barriers (but not elimination of the EU common external tariff). Elimination of non-tariff barriers was found likely to increase aggregate trade volumes in all industries. Estimates of increase ranged from 16 per cent for pharmaceuticals to 133 per cent for electrical machinery. In every industry, the percentage increase in intra-EU trade volume was estimated to be larger than the aggregate, suggesting some displacement of imports from non-EU countries. In tonne-kilometre or truck kilometre terms, the effect of eliminating non tariff barriers is to increase international road haulage by 38 per cent. The results confirm the expectations of an EU Task Force, which had predicted that shifts toward liberalization and deregulation might increase trans-frontier truck traffic by 3050 per cent (CEC 1989).4
The scope for reducing freight volumes
There are many different feasible mechanisms by which to reduce the environmental impact of freight transport. The focus here is on the potential to reduce emissions by reducing the transport intensity of products (mostly through changes in the spatial division of labour) and by increasing the efficiency of use of existing modes - both within the context of existing infrastructural constraints.
Reducing the transport intensity of products
The scope for reducing the transport intensity of final products lies in several possible adjustments. One is to reduce the number of transport-intensive components that products embody. Another is to reduce the average length of individual transport movements. This can be done by obtaining raw materials and components locally, serving local markets, reversing current trends in logistics, and revising networking arrangements in respect to contractors and subcontractors in the value-adding chain. Finally, there is also scope to reduce the number of times that the same materials are transported during the process of adding value.
Reductions in the number of transport-intensive components and ingredients can be achieved by substitution. This is particularly im portent in the food industry because of the high quantities of food products now involved in long-distance transport and the high level of substitutability among foods and ingredients. This would imply greater reliance on local/regional produce and on produce in season.
Some physical goods movement can be substituted altogether by using the possibilities provided by flexible technologies and informatics. The possibility exists for information about the design of components embodied within complex products that are manufactured at a distance from the market to be provided with the product or transmitted electronically as need arises. In the event, say, of the product breaking down, a local parts manufacturer is thus able to obtain component design details and make a replacement part rather than having to order and ship this in from a distance.
Similarly, telecommunications can substitute for information-rich but formerly transport-intensive products and services. Just as fax and electronic mail can substitute for conventional mail, similar substitutions can be made in respect to whole industrial sectors. In the newspaper publications industry, for example, a combination of telecommunications and flexible printing plant could do away completely with the need to transport newsprint physically over long distances. It also provides a potentially improved customer service in the form of a greater choice of simultaneously published editions.
The long-distance shipping of goods between world regions could increasingly be replaced by capital flows by encouraging direct foreign investment as an alternative to trade. The attractiveness of direct foreign investment increases with the level of competition in non-standardized products and with the rapidity of market change. Moves toward the strengthening of regional trading blocs (such as the KU) could help in this direction.
Changing the spatial structure of inter-firm supply networks (the number and relative location of contractors and subcontractors) is important because these needlessly increase the transport intensity of final products. Strutynski (1995) argues that such traffic-generating arrangements are currently promoted by EU programmes aimed at integrating European economies, technology transfer, and increasing international marketing opportunities. Only international projects are promoted. Thus he says that: "It is easier to establish co-operation between enterprises in Northern Hesse and the North of Spain than between two neighboring enterprises, for example in Baunatal and Kassel" (Strutynski 1995, p. 39).
Strutynski argues that there is scope for lean production to con tribute to reductions in inter-firm freight transport if reductions in the number of direct suppliers are accompanied by reductions also in average transport distances between them. This depends upon regional concentration of production and supply relations. In a theoretical example based upon German conditions, he contrasts the transport implications of two different inter-firm supply structures. Both are geared toward producing an identical final product. The first represents a traditional relationship between final assembler and several suppliers in which all suppliers deliver directly to the assembler. The second represents a relationship based upon cooperation among suppliers organized into regional value-adding networks. The effect is that more value is added within the region before any components are transported long-haul. Only a small number of high value-density components are finally transported to the assembler. In his example, Strutynski (1995) demonstrates a theoretical 70 per cent reduction potential in total supply distance.
Increasing the efficiency of use of existing modes
Average loading ratios for trucks in Europe are typically 75-85 per cent (ECMT 1995). Empty truck movements account for approximately 1.5 per cent of all domestic freight traffic within countries of the KU. Prior to deregulation and the creation of the Single Market in 1992, one-third of trucks making international journeys on EU roads had empty back-hauls. Of trucks crossing the Dutch-German border, 30 per cent were empty (Gabel 1994). Although all these data are for the immediate pre-1992 situation, moves to phase out permits and restrictions on cabotage post-1992 were not anticipated to improve loading ratios. Surveys of hauliers undertaken before 1992 show the main causes of low loading factors to be related not to restrictions but rather to problems in arranging suitable backloads for example, a lack of information on backload opportunities and problems of backload incompatibility (CEC 1989). Broadly similar loading ratios are likely to apply currently.
The problem of low loading factors is linked to operators' incentives. Some operators prefer to operate large vehicles because this maximizes flexibility and reduces average costs. As size restrictions have been relaxed, the average truck size of some operators' fleets has increased.5 Once equipped with large vehicles, hauliers use these even with small loads (Zuckerman 1991).6 The difference between actual and theoretically feasible loading factors implies scope for efficiency gains equivalent to 25 per cent of total kilometrage. While this may not be realizable owing to practical problems of specialization and backloading, these constraints are themselves related to the overall paradigm of which scale economies, long-haul transport, and specialization of transport are part. Packaging is also part of this paradigm and part of the problem. Goods for final sale are transported together with often bulky packaging. This reduces weight-to-volume ratios and increases truck-kilometres. The scope for efficiency gains is therefore a function not only of truck loading ratios but also of weight-volume ratios.