The Economics of Urban Transportation

Kenneth A. Small, Erik T. Verhoef

• Publisher: Routledge
• Year: 2007
• ISBN:
  9781134495702
• Edition: 2
• Topic: Business
• Subtopic: Business General
• Index: Business

This timely new edition of Kenneth A. Small's seminal textbook Urban Transportation Economics, co-authored with Erik T. Verhoef, has been fully updated, covering new areas such as parking policies, reliability of travel times, and the privatization of transportation services, as well as updated treatments of congestion modelling, environmental costs, and transit subsidies.

Rigorous in approach and making use of real-world data and econometric techniques, it contains case studies from a range of countries including congestion charging in Norway, Singapore and the UK, light rail in the Netherlands and freeway tolls in the US.

Small and Verhoef cover all basic topics needed for any application of economics to transportation:

• 
  
  

• forecasting the demand for transportation services under alternative policies
• measuring all the costs including those incurred by users
• setting prices under practical constraints
• choosing and evaluating investments in basic facilities
• designing ways in which the private and public sectors interact to provide services.

This book will be of great interest to students with basic calculus and some knowledge of economic theory who are engaged with transportation economics, planning and, or engineering, travel demand analysis, and many related fields. It will also be essential reading for researchers in any aspect of urban transportation.

1 Introduction

At the heart of all modern economic activity is trade. People trade labor and ideas for cash, and cash for goods and services; firms trade technology, expertise, financial capacity, intermediate goods, administrative functions, and many other things with each other, with individuals, and with governments. All these transactions require communications and most require transportation of goods or people – to work, shopping, tourist sites, meeting locations. Thus it is fair to say that transportation is central to economic activity.

Cities exist when there are special advantages to carrying out economic activities in proximity, advantages often called “economies of agglomeration” because costs are lower when certain groups of activities locate close to each other. The primary reason for agglomeration economies, especially in a world with low communication costs, is that transportation costs are still significant and proximity reduces them. A corollary is that anything that reduces transportation costs within an urban area increases the extent to which its activities are easily linked to each other, and thus takes further advantage of agglomeration economies. In a world of many competing urban centers, those with more efficient transportation systems have an advantage.

The study of transportation involves researchers trained in many disciplines, including engineering, economics, geography, planning, psychology, business, and regional science. Regardless of its disciplinary origin, transportation research has become increasingly sophisticated in its use of economics. This trend has brought a solid practical footing to policy analysis, by showing how the ideas and goals generated within various analytical frameworks can be reconciled with the actual behavior of users of transportation systems and with the resource constraints of a real economy.

This book reviews the contributions that economics can make to the analysis of urban transportation. It concentrates mainly on industrialized nations, but many of the principles are equally applicable to developing nations. This is especially true in light of the ubiquitous rise in travel by private automobile. Because of the predominance of this trend, we place heavy emphasis on highway transportation and use it as a topic in which to present a comprehensive set of formal models.

The scope of urban transportation economics

The boundaries of transportation economics are neither well defined nor static. Nevertheless, the reader is entitled to know what principles we use to limit the scope of this review. Aside from the inevitable one that we try to write about what we know best, the following observations are guidelines.

Transportation economics is, of course, a branch of economics. Hence it focuses on resource allocation and how the interactions among independent agents bring about a self-consistent outcome. It draws from and interacts closely with transportation engineering, urban planning, and other disciplines, but has a somewhat different emphasis. Engineering emphasizes facility design and implementation, while economics emphasizes behavioral principles and resource allocation. The disciplines of management, public administration, and urban planning are concerned with the formulation of workable transportation policies, for example by studying decision processes and organizational structures. An important role of economics is to inform these disciplines about the complex ways in which transportation policies exert their influence. Economics is well suited to predict the ultimate results of behavioral shifts among interacting economic actors in response to policy implementations or exogenous events. It also can identify tradeoffs between efficiency and other goals.

These orientations toward the subject give it certain characteristics that are evident in this review. Transportation economics tends to focus on models that illustrate concepts, as opposed to those whose output is the actual design of facilities or regulations. Hence its models are usually at a coarse rather than a fine geographical scale, lending themselves to “sketch planning” of the broad features of a transportation system.

Analysis often proceeds by defining a demand structure and a supply structure for a set of goods or services, then searching for an outcome that is consistent with both structures. This is a normal microeconomic approach, although the nature of transportation creates ambiguities in the boundary between demand and supply: for example, is the time required for a trip an attribute affecting demand, or is it part of the cost? Either viewpoint is valid so long as demand and cost are consistently defined; doing so is the task of Chapters 2 and 3, respectively. Their interconnection is made explicit in the discussion of the “value of time” in Section 2.6.

Demand and supply structures are complex, involving many types of people, modes, locations, and times. For this reason, finding a consistent solution–an equilibrium, in economics parlance–requires considerable analytical sophistication. A pioneering study by Beckmann, McGuire and Winsten (1956) showed how to do this for arbitrary transportation networks given certain assumptions. In the presence of increasing returns to scale and lack of efficient pricing (conditions which, as we shall see, often characterize urban transportation), this equilibrium may differ in drastic and surprising ways from a configuration satisfying optimality conditions. In other words, what happens when people make decisions out of self-interest is not necessarily what they would choose collectively and cooperatively. A common thread throughout this review is that transportation economics should provide the tools to understand and quantify such differences, and to design policies that address them.

Three important types of policies to consider are pricing, investment, and industrial organization (including regulations contingent on the form of organization). Pricing and investment have long been the hallmark of urban transportation economics, and are accorded full treatment here (in Chapters 4 and 5). Industrial organization has likewise been of great concern, and recent changes throughout the world have brought to the fore questions of oligopolistic markets, regulation, and government versus private provision of services. Chapter 6 considers a number of these questions.

Transportation potentially affects the nature of the urban area itself. If transportation were costless, participants in an economy would have no economic reason to locate close to one another. The study of this influence is clearly germane to transportation policy. To analyze it fully requires the full power of disciplines such as urban geography, urban economics, and regional science, which seek to explain the shape of urban development. Important as this is, it is too ambitious for this book and we limit ourselves to a brief summary in Section 2.1.6.

The scope of this book

The aim of this book is to provide a self-contained introduction to major research themes in contemporary transportation economics. We have chosen to explore certain topics in great detail, in order to illustrate how the analysis can be put to practical use, and others much more cursorily. In both cases, extensive citations provide the reader opportunities to expand on any particular theme. Thus we aim to make the book suitable both as an initial textbook for students with a good technical background and as a reference work for practicing researchers and professionals.

Above all, we attempt to show how to construct a set of workable models that can be adapted, refined, and combined in subsequent research. In the sections on highway congestion, for example, the models presented incorporate critical features that are often omitted in transportation analyses, such as queuing, trip scheduling, peak shifting, and reliability. There is evidence that these phenomena greatly affect the outcomes from policies to relieve congestion, and their analysis has been made tractable by recent technical advances. It is our hope that such models will provide a common language for researchers in urban transportation, facilitating comparisons among theoretical innovations and among empirical applications.

2 Travel demand

In order to plan transportation facilities, it is necessary to forecast how much they will be used. In order to price them rationally and determine the best operating policies, it is necessary to know how users respond to prices and service characteristics. In order to evaluate whether a project is worthwhile at all, it is necessary to have a measure of the benefits it produces. All these requirements are in the province of travel demand analysis.

The demand for travel takes place in a multi-dimensional setting. The traditional sequential framework used by many metropolitan transportation planning agencies considers four choice dimensions: trip generation (the total number of trips originating from an area); trip distribution (the locations of the trips’ destinations); modal choice (the means of travel, such as car, bus, train, bicycle, or walking); and trip assignment (the exact route used). More recently, researchers have paid greater attention to other dimensions of choice, such as residential and job location, household automobile ownership, the time of day at which trips are taken, parking locations, and the duration of activities for which travel is undertaken.

These multiple decisions are often envisioned as a sequence, typically starting with residential and job locations, then vehicle ownership, then other aspects. This sequence is in decreasing order of the time span over which the decision can be changed easily. However, it does not imply a sequential decision procedure whereby one decision is made without regard to its implications for later decisions. Rather, each decision is affected by others and so can be fully understood only as part of a simultaneous choice process. A given study may isolate just a few of these decisions for tractability; it is then all the more important to remember, in interpreting results, that other decisions are lurking in the background.

Furthermore, travel is a derived demand, usually undertaken not for its own sake but rather to facilitate a spatially varied set of activities such as work, recreation, shopping, and home life. This observation links the study of travel demand to studies of labor supply, firms’ choices of technologies, and urban development. It also calls attention to an increasingly common form of travel: the linking together of several trip purposes into one integrated itinerary or tour, a process known as trip chaining.

The chapter begins (Section 2.1) by asking what can be learned from aggregate data about travel and how conventional economic demand theory is applied to such data. It then moves on to disaggregate models (Section 2.2), also known as “behavioral” because they depict individual decision-making explicitly. Section 2.3 presents examples of models explaining some key travel choices: mode, time of day, and route. More specialized topics are then discussed (Sections 2.4–2.5). Finally, Section 2.6 analyzes two quantities of special interest to policy: travelers’ willingness to pay for travel-time savings and for improved reliability. We mostly discuss passenger transportation, in part because more data are available than for freight; however, studies of the demand for urban freight transportation tend to use similar methods.¹

2.1 Aggregate tabulations and models

Much can be learned simply from cross-tabulating survey data. For the United States, there are two very useful sources. One, covering all trips, is the National Personal Transportation Survey (NPTS), collected at approximately six-year intervals; for the single year 2001 it was subsumed into a broader survey called the National Household Travel Survey (NHTS). The other useful source is the journey-to-work portion of the US Census, taken every 10 years. This consists of responses to questions about work travel, asked on the Census “long form,” which is administered to about 17% of all households. Pisarski (2006) provides a comprehensive analysis of the 2000 journey-to-work census and compares it to earlier censuses and to the 2001 NHTS.

The NHTS shows that work trips in the US continue to decline as a proportion of all trips, amounting to about 16% in 2001.² However, it would be a mistake to conclude that work trips are no longer important for urban transportation. Two-thirds of US work trips in 2000 occurred during the hours 6–9 a.m., and work trips account for about half of all person-miles of travel during those hours. Trips to work have been found to account for more than half of all trips in selected Belgian cities, and a very high fraction of peak-period trips on some of the most notoriously congested Los Angeles freeways.³ Thus work trips remain predominant contributors to congestion.

Private vehicles dominate all categories of trips in the US, and increasingly so. They account for 91% of all work trips in 2000, up from 89% in 1990. The share of work trips that use public transit is just 4.7%, down from 5.3% in 1990. However, transit share is much larger in large metropolitan areas, averaging 5.7% and reaching 11.5% in those areas with over 5 million residents. Carpooling accounts for 12.6% of all metropolitan work trips. Automobile ownership is high and continues to grow faster than the population. Yet many US households owned no car in 2001: roughly 8% of all households and 26% of those with incomes below $20,000.⁴

The geography of c...