Network Effects and Total Economic Impact in An Enlarged Trans-european Transport Network
J Laird, J Nellthorp, P Mackie, ITS, University of Leeds, UK
There is a widely held belief among policy makers that network effects are real and will be an important source of benefit from the enlargement and enhancement of the Trans-European transport network (TEN-T). There is also a perception that these network-derived benefits are additional to those that would be captured through a normal economic appraisal. The enlargement of the TEN-T network is a cornerstone of the European Commission?s transport policy set out in the White Paper ?Time to Decide? (EC, 2001), and many of the initiatives specified there are fundamentally about developing ?network? aspects to the TEN-T. For example, harmonisation of track gauge between Spain and Portugal and the rest of the EU will create an interoperable High Speed Rail Network, whilst the initiatives to improve border crossings between the EU15 countries and the newly associated states will create a more cohesive transport network within the enlarged EU. Policymakers therefore require clear guidance on the analysis of network effects and their inclusion in the economic impact appraisal. This paper provides a basis for that guidance, as well as identifying the areas where future research is needed. It draws on the network economics literature to understand the causes of network effects, as well as research conducted within the 5th Framework project IASON (2001-4).
The recent literature on network effects within the transport sector is, surprisingly, not extensive. The most notable work has been to produce a nomenclature for network effects and some basic empirical results (Roy, 1995; Turro, 1999; Van Exel et al, 2001; Mackie et al, 2001). Roy (1995) introduced the concept of European Value Added (EVA) as a means of describing the benefits that accrue to countries other than the one in which the transport project is located?an issue of key interest to the EU?but there is more to network effects than this.
There exists an inter-disciplinary literature on network effects involving the disciplines of mathematics, economics (particularly the treatment of electricity, gas and telecommunication networks) and sociology, which provides a base upon which analogies can be drawn for the treatment of network effects within transport (see, for example, Katz and Shapiro, 1994; Economides, 1996; Shy, 2001, cited in Mackie et al, 2001). Within the transport field there is also a body of literature on links between transport and the wider economy (for example SACTRA, 1999; OEEI, 2000), which has a ?network effects? dimension.
Networks are in simple terms nothing more than a system, comprising objects and connections between those objects (Mackie et al, 2001). Among the features which make transport networks interesting from an economic point of view are the presence of: (i) economies of scale, scope and density in the supply of a network?espcially public transport networks (e.g. Winston, 1985; Jara-Diaz and Basso, 2003), and (ii) congestion when demand interacts with supply. The implications for transport policy are that when networks are dramatically expanded (e.g. Iberian High Speed Rail integration; east-west European rail and roads) or when networks are restructured (e.g. airlines? hubs and spokes) or when traffic density is increased, there may be spin-off benefits to all network users through a reduction in network costs per traffic unit. Conversely, any policy-driven stimulus to transport demand which drives traffic onto congested networks may have the opposite effect of external costs to other network users. These positive and negative transport network effects are a perennial issue for national and EU transport planners, however, IASON has found that they are treated only partially in most current EU-level transportation models (Laird et al, 2003).
Network effects can also occur within the wider economy as a result of a transport initiative (Mackie et al, 2001). In this case, the effects are due to the presence of consumption externalities, agglomeration economies and economies of scale in production (e.g. Venables and Gasiorek, 1998; Oosterhaven and Elhorst, 2003). Such conditions could be expected to occur, for example, in remote or peripheral regions of the EU that may be subject to a degree of monopolistic competition and labour markets with excess supply. Network effects in the wider economy may therefore prove to be an important source of benefit for TEN-T network projects that involve improving the cohesion of the EU, and particularly with regard to links to the new member states.
Network effects can only give rise to additional economic impact benefits if the appraisal is based on an inadequate transport-economy model, however the international scale of the TEN-T makes this potentially likely. Truncating the model area to national boundaries, the use of coarse zoning systems, inferior representation of public transport networks in EU-wide models, and omission of feedbacks between the wider economy and transport sector and between demand and supply are all potential sources of inadequacy (Laird et al, 2003). For traditional transport project appraisals, the lack of a link to a model of the wider economy places a considerable burden on the transport model elasticities. Early empirical results on the inclusion of network effects in the appraisal of the TEN-T are contradictory?some show large positive additionality, some do not. For the full paper, we expect to present more definite empirical results on this matter.
Association for European Transport