قیمت گذاری بهینه اثرات جانبی حمل و نقل در محیط بین المللی: برخی نتایج تجربی بر اساس یک مدل بهینه سازی عددی

In small open economies, a large fraction of the externalities on the national transport network is due to international and pure transit traffic. Moreover, the existence of these flows implies a potential for tax exporting behaviour by individual countries. This paper provides an empirical analysis of welfare-optimal pricing of transport externalities in such an environment. A numerical optimization model is used to study different types of equilibria. We consider the welfare optimum for one individual country (Belgium) assuming that current policies abroad (in ‘the rest of Europe’) remain unchanged, and we compare the outcome with current pricing policies and with the welfare optimum from the viewpoint of a hypothetical federation, consisting of Belgium and the rest of Europe. We also provide some first insights into the strategic responses of individual countries to transport policies abroad by numerically evaluating the Nash equilibrium in transport prices. Among others, the empirical results indicate the order of magnitude of deviations of current prices from optimal prices, they numerically illustrate the importance of tax exporting behavior, and they provide information on the response of individual countries to changing transport policies abroad.

The large increase in mobility over the past decades has generated serious concerns about the associated external costs such as congestion, noise, accident risks and air pollution. Not surprisingly, a substantial economic literature has developed that deals with optimal pricing and regulation of these transport externalities. The theoretical literature on optimal taxation in the presence of externalities (see, e.g., Sandmo 1975; Bovenberg and van der Ploeg, 1994) has been extended to specifically fit the characteristics of the transport industry (Oum and Thretheway, 1988 and Mayeres and Proost, 1997). Moreover, a number of studies have considered the pricing of transport services under various second-best conditions (see, e.g., Arnott et al., 1993, Verhoef, 1996, Parry and Bento, 2001 and Small and Yan, 2001). An impressive stream of empirical studies is now also available in which the pricing implications of congestion and other external costs were numerically illustrated. Relevant references include Glaister and Lewis (1978), Small (1983), Viton (1983), Kraus (1989), De Borger et al. (1997), and Proost and Van Dender (2001). The majority of the above-mentioned studies concentrated on correcting externalities in an urban environment or on a specific road network, and has therefore not explicitly considered the international aspects of the problem. It is clear, however, that transport flows and the associated externalities have an important international dimension that must be taken into account in designing realistic pricing and regulatory policies. First, the importance of international (and especially pure transit) transport flows implies that in many countries a substantial share of locally generated externalities is caused by foreigners. Second, to the extent that countries can tax these flows, tax-exporting behavior may result (Arnott and Grieson, 1981 and Krelove, 1992). Third, international transport implies that the tax base of transport services is to a large extent mobile between countries, which may lead to inefficient tax competition. Indeed, increases in tax rates in a given country imply a fiscal externality that is likely to be ignored by individual countries (see, e.g., Mintz and Tulkens, 1986 and Wildasin, 1988). Fourth, some transport externalities generate international spill-overs (e.g., global warming, acid rain, etc.). These will lead countries to underestimate the global effects of locally generated externalities and to impose corrective taxes that are too low from a global perspective (see, e.g., Markusen, 1975 and Merrifield, 1988). As far as we know, this paper provides the first empirical analysis of optimal pricing of transport externalities in an international environment. The theoretical framework we use captures all essential dimensions of transport policy-making in an international setting, including spillovers of externalities and the potential for tax competition and tax exporting behavior. However, we do make some strong simplifying assumptions to keep the problem empirically manageable. Indeed, since in a realistic international environment (e.g.,, within the EU) countries experience transport from a large number of foreign countries on their territory, a fully satisfactory empirical model would have to include many individual countries and all inter-country traffic flows. In this paper we strongly reduce the data requirements and computational complexity by focusing on two asymmetric regions, Belgium and its surrounding region, the ‘Rest of Europe’. In each region we distinguish passenger and freight flows. Whereas passenger transport is assumed to be purely domestic, freight transport in a given region includes both domestic and international flows. The model allows for many transport modes and services and incorporates, although in a highly stylized way, general equilibrium effects of freight transport prices on other goods in the economy. Each region can tax all passenger and freight transport flows occurring within its jurisdiction; consistent with current international legislation (see, e.g., EU legislation), no discrimination between domestic and international freight flows is allowed. The empirical analysis considers three stylized types of policies. (i) We look at the welfare optimal tax policies from the perspective of one individual country, conditional on existing transport policies elsewhere. This illustrates the best a country like Belgium can do if it takes current tax policies in the rest of Europe as given and if it is interested in accounting for external costs imposed by both domestic and international transport flows on its territory. (ii) We consider optimal transport policies from the viewpoint of the hypothetical federation consisting of Belgium and the rest of Europe. This exercise provides information on optimal policies from the perspective of a European ‘government’ that takes account of external costs and of all interactions between countries, both in terms of transport flows and in terms of tax revenues. (iii) We provide a first, and admittedly crude, assessment of strategic behavior between countries by looking at the Nash equilibrium. To guarantee that the non-cooperative two-region equilibrium would mimic as closely as possible the outcome of the game involving all 12 individual countries, this exercise tries to avoid two unrealistic consequences of aggregating countries in a strategic setting. Indeed, treating the rest of Europe as one homogeneous entity implies that all spill-overs of externalities between members of the aggregate country would be automatically internalized. Moreover, it would imply perfectly coordinated policy actions by all the member countries of the aggregate. The correction procedure, based on Nordhaus and Yang (1996), circumvents these unrealistic features while still allowing us to derive interesting results in a two-country framework. However, it is itself based on some very strong assumptions so that the Nash equilibrium should be interpreted as a crude and modest first attempt at identifying strategic interactions of transport policies between countries. The results of the paper provide numerical evidence on at least three issues. First, what is, from the viewpoint of the federation, the order of magnitude of the necessary tax and price adjustments to cope with the external costs transport services generate, taking into account all international dimensions of the problem? Second, to what extent does the presence of both domestic and international freight transport (or pure transit) induces countries to engage in tax exporting behavior? Third, what is the order of magnitude of the difference between federally optimal prices and the tax structure that results if countries compete strategically? The paper proceeds as follows. In Section 2 we provide a description of the theoretical structure of the model, and we describe in detail the implementation of the numerical optimization model. Section 3 provides information on the data used and reports on the empirical results. It consecutively analyses the characteristics of the reference situation, the federal optimum, the local optimum from the point of view of one individual country, and the Nash equilibrium. Moreover, it suggests answers to the three questions raised above. Section 4 concludes.

In this paper we studied optimal pricing and regulation of transport externalities in a federation. The analysis illustrated the role of both domestic and international transport, of external cost spillovers, and of tax exporting behavior associated with the prevalence of international flows on country’s domestic network. A numerical optimization model was used to determine optimal policies for a federation consisting of two asymmetric countries, Belgium and ‘the rest of Europe’. The application compared four types of equilibria: the reference situation reflecting unchanged policies, the federal optimum, the local optimum for one individual country, and the Nash equilibrium solution. The empirical application provided evidence on the required price and tax adjustments on a large number of different sub-markets, defined on the basis of transport mode, period of the day, type of car, etc. With the exception of the reference situation, taking into account the external costs generated by transport flows results in higher taxes on a large variety of transport services for all other equilibria considered. At the federal optimum, it was shown that optimal pricing to cope with externalities requires substantial price increases of the majority of transport services, especially peak road transport. Similar price increases were observed at other equilibria. However, both the local optimum for Belgium and the Nash equilibrium demonstrated the inefficiency of tax exporting behaviour: taxes on freight flows were found to be substantially higher than in the federal optimum, depending on the importance of international flows on the domestic territory. This yielded an overall decline in welfare for the federation as a whole. The local optimum for Belgium involved higher welfare than both the federal optimum and the Nash equilibrium outcome. The former was due to tax exporting, the latter to the absence of reactions by the rest of Europe.