This article examines the role that economics can play in analysing problems with urban transportation in the United States. The specific problems addressed are failing infrastructure, financially weak public transit, environmental impacts of motor vehicles, motor-vehicle accidents, and traffic congestion. Simple quantitative analyses, even though approximate, can help to focus attention on the most promising classes of policies. Those classes involve some technological measures and some narrowly targeted behavioral changes, but not the widespread curtailment of of motor vehicles use.

When a firm or public authority prices output at marginal cost, its profits are related to the degree of local economies of scale in its cost function. As is well known, this result extends to the case where some congestion-prone inputs are supplied by users. I show that contrary to common belief, the result holds even when scale economies are affected by a rising factor supply curve. In that case, constant returns to scale in production produces diseconomies of scale in the cost function, making marginal-cost pricing profitable. Examples are provided for a monopsonist both with and without price discrimination. In the latter case, second-best pricing is also considered: profits then are not governed in the usual way either by returns to scale in production or by scale economies in the cost function, but some useful bounds are provided.

Transportation policy making frequently requires evaluating a proposed change, whether it be a physical investment or a new set of operating rules for allocating rights to an existing facility. Some, like the rail tunnel under the English channel, are one-time capital investments with enormous and complex effects on accessibility throughout a network. Others, like congestion pricing proposed for Hong Kong, may be technically reversible but require major behavioral and political groundwork. 

In such cases, the optimization framework that proves useful in so much transportation analysis is often inadequate. In an optimization model, important aspects of a problem are represented as a few variables which can be chosen to maximize some objective. For example, Robert Strotz shows how highway capacity can be chosen to minimize total travel costs in the presence of traffic congestion. But often the change is too sharp a break from existing practice, or the objectives too numerous, to represent the problem in a mathematical optimization framework. Perhaps a given highway improvement not only expands capacity to handle peak traffic flows but also speeds off-peak travel, reduces accidents, and imposes noise on residential neighborhoods. Perhaps the required capital expenditures occur in a complex time pattern, and the safety effects depend on future but uncertain demographic shifts. One would like a method for analyzing the merits of such a package of changes, and for comparing it to alternative packages. 

Such a method is called project evaluation. Performed skillfully, it can identify key consequences of a proposed project and provide quantitative information about them to guide policy makers. Much of this information may be non-commensurable: i.e., the consequences may not all be measured in the same units and hence the analyst may not be able to determine the precise extent to which these effects offset each other. For example, a tax-financed improvement in airway control equipment might improve safety but magnify existing income inequalities. 

Methanol is widely regarded as a promising alternative to petroleum-based fuels in a wide variety of uses, including transportation. The primary advantages would be lower air · pollution and diversification of energy sources. Interest is strong in California and especially in the Los Angeles air basin, leading to recently announced plans by the South Coast Air Quality Management District to begin converting large fleets of cars and buses to methanol. 

Transit buses are an especially promising place to begin a strategy of using methanol as a transportation fuel. Their emissions are very visible and affect crowds of people, and the buses themselves are mostly operated in fleets by public agencies. In addition, federal emissions standards for heavy-duty diesel vehicles are especially strict for transit buses starting in 1991. 

Our project focused primarily on the air pollution benefits of converting transit buses to methanol fuel. We used two different methods: cost-benefits analysis and cost-effectiveness analysis. Most of the analysis was carried out at one assumed price of diesel fuel and one or more assumed prices for methanol fuel, but we have also considered the mechanisms by which the two prices might be linked together.