Positive Externalities and the Public Provision of Transportation Infrastructure: An Evolutionary Perspective ULRICH BLUM Dresden University of Technology ABSTRACT Do transportation systems, comprising infrastruc- ture, service, and use, produce external benefits? If they do, should positive externalities be accounted for in the evaluation of infrastructure investments? This paper argues that while direct, technological, external benefits from transportation are difficult to find, meaningful positive externalities can arise from transportation systems in at least two ways. First, transportation infrastructure can reduce pre- existing negative externalities, and the reduction of external cost must be considered an external bene- fit. Second, because transportation is essentially a derived demand its effects are broadly diffused throughout the primary markets that induce trans- portation demand. To the extent that changes in transportation infrastructure induce positive exter- nalities in these primary markets, external benefits should be attributed to transportation. A ONE-SIDED EXTERNALITIES DEBATE? The general discussion on externalities of trans- portation—be they monetary or technological— usually concentrates on negative effects. In this paper, we concentrate on positive effects, although Professor Dr. Ulrich Blum, Chair of Economic Policy Re- search, Dresden University of Technology, D-01062, Dresden, Germany. Email: [email protected]. 81 most scholars would question their existence—and external benefits (e.g., the beauty of a bridge in the in most cases they are right if the external effect is case of infrastructure) or something comparable to defined purely on a technological basis, leaving out “positive emissions.” The very abstract concept of monetary effects. mental maps derived from a high degree of mobil- Consequently, most of the approaches used to ity and facilitating extended interaction spaces evaluate the effects of externalities relate to the dif- might be useful, though very unlikely. ficulty of pricing negative technological external The question is: what is ground zero? To what effects.1 We may distinguish between: sort of alternative do we compare the existing m The resource approach. The value of the exter- transportation system? If we know the reference nality is defined by the corresponding resource point, all effects “below” may be referred to as price of the private market, which in most cases negative externalities, all effects “above” as posi- relates to prices for damage or repair. tive externalities. Of course we know that this m The avoidance approach. The value of the exter- objective reference point does not exist, and there nality is defined by the possibility of substituting is no chance to objectively derive it. Thus, reduc- the resource, the technology, or the good in tions in negative externalities may be considered as question for a resource, technology, or good net positive externalities. without the external properties. A more challenging issue seems to be that trans- m The risk approach. The value of the externality portation demand is derived from other markets. If is defined by the discounted expected monetary external effects exist in these “primary markets,” value based on an evaluation of risk. they are in a causal sense transferred from one m The utility approach. The value of the external- place to another. This link is not restricted to geo- ity is defined by the willingness to pay in order graphical locations; it may also refer to other met- to reduce negative effects. rics, that is, product space. The capacity of these The scientific argument behind this apparent links may be an important factor in the transfer of one-sidedness—which only accounts for negative externalities. At the limit, without the transfer no effects—is that the positive effects of transporta- market would exist and, thus, no externality issue. tion investments are immediately captured by mar- In this paper, we investigate the external benefits kets, unless market failure impedes it, whereas of transportation stemming from spatial transfers most negative effects remain external. This view is from one primary market to another or from sup- reflected in most of the literature.2 Policy measures ply to demand. From what is known about spatial not correcting for this asymmetry must necessarily analysis (Blum 1996), it is clear that once markets lead to allocation failures. or points of production and consumption are The most intuitive examples of negative exter- linked by means of interaction (i.e., transportation nalities might be the adverse effects of infrastruc- in geographical markets, substitutive exchange in ture, for example, dividing up a landscape or product markets), the derived transportation vehicle emissions. It is very difficult to imagine demand (or transfer) function reflects excess demand and supply as well as underlying external- ities. Furthermore, the effect will depend on the 1 The willingness to pay for reduced noise normally differs from costs for technological improvements in cars or actu- degree of perfection of these markets, that is, the al costs incurred in the medical sector. more they are monopolized, the smaller the trans- 2 The analysis of negative externalities of transportation in fer effect. Germany reveals costs that vary between 44 bn DM and 221 bn DM p.a. (IWD 1996). Growth effects are mostly Let us look at some intriguing examples: analyzed within the framework of production models m A vaccine prevents the spread of a disease once (Blum 1982). External benefits have been reported in a sufficient proportion (below 100%) of a pop- Germany by Aberle (1992, 1994), who especially stressed ulation has been vaccinated. Thus, all those who the spatial aspects, the aspects of market integration, and cost savings; however, it has been argued that these bene- are not vaccinated benefit from an external ben- fits are not external at all as they are included in the allo- efit. If such a vaccine is shipped into a new mar- cation via the price system. ket, external benefits emerge there, and their 82 JOURNAL OF TRANSPORTATION AND STATISTICS OCTOBER 1998 level will depend on demand that, again, will be sumption of such a good leads to no subtraction partly influenced by transportation costs— from any other individual’s consumption of that which makes the institutional setting of the good.” As a consequence, consumers cannot be transportation market of crucial importance. excluded from consumption and there exists no m Knowledge produced in one region is trans- rivalry for consumption. ferred to another region by means of commerce: In this definition, the pure public goods are one the level of external benefits in the sense of pole of a continuum that extends to pure private spontaneous adjustments to the economy will goods at the other end. It was quickly discovered, depend on the intensity of this transfer, and it however, that ideal public goods rarely exist in will be difficult to capture these effects from the reality and that private allocation is possible once beginning because of the existence of non- mechanisms can be found that: knowledge or uncertainty (Knight 1921). m prevent the free use of these public goods, and Market integration, however, triggered by the m force groups to reveal their preferences, at trans- same transport system may produce sufficient action costs that do not reduce demand to zero. information to grasp the issue and internalize it. Symmetrically, pure private goods are also only m The existence of a transportation system allows an ideal counterpart to the ideal public good; the production and consumption nodes to interact, spectrum between these poles may be seen as a enabling them to produce economies of scale continuum that we will address later. This leaves and of scope, and network economies. Accord- unresolved the question of what type of good a ing to the new growth theory, they are external road really is, for instance. In fact, it is easy to find to the firm and can only be captured once suffi- roads or other types of infrastructure that can be cient information is available. positioned at different points on this continuum. m Visitors to a resort island benefit from the beau- The following observations may serve as a first ty of nature, depending on the number of peo- challenge to the easy categorization of infrastruc- ple allowed on the island. ture: All these external benefits are surpluses in the m there is no such thing as rivalry per se; it de- transportation demand system; in the price-volume pends on congestion and thus may vary widely diagram, a social transportation demand function for the same infrastructure. Once a motorway is with a higher reservation price reflects the level of filled to capacity, the situation is different from demand if all positive externalities were properly when it was empty. accounted for by the individuals. m exclusion depends on the relative weight of The structure of the rest of this paper is as fol- transaction costs that vary in space and time. lows: the next part classifies goods according to Some of these transaction costs are “natural,” demand and supply characteristics; then, a model while some are created in order to force the rev- is proposed to link the characteristics of goods to elation of preferences. varying congestion levels, a dominant characteris- m the quality of the public good may depend on tic of the use of infrastructure. Finally, we show location, that is, a constant quality and/or quan- that this classification provides new insights into tity is not guaranteed.3 the way infrastructure allocation is organized and In fact, we see that given a fixed infrastructure costs and benefits are evaluated. capacity, exclusion reduces congestion and there- fore also rivalry. INFRASTRUCTURE, PUBLIC GOODS, Our main argument is as follows: the quality of AND EXTERNALITIES a public good is not only a supply-side characteris- tic but also relates to demand (i.e., preferences).