Reviews that appear in JTS describe and assess books about new developments in statistics, economics, the environment, or engineering research €that focus on transportation issues. The topics can be theoretical, empirical applications, or methodological innovations.
Suggestions of books for review are welcomed, as are requests to become a book reviewer. Please contact the Book Review Editor:
Prof. Vincent W. Yao
2801 S. University Avenue
Institute for Economic Advancement
University of Arkansas at Little Rock
Little Rock, AR 7204-1099
(501) 569-8453; firstname.lastname@example.org
Transportation Labor Issues and Regulatory Reform
James Peoples and Wayne Talley, editors
November 2004, 179 pages
$95 € 86 £57.50
Transportation Labor Issues and Regulatory Reform, volume 8 in Elsevier's Research in Transportation Economics series, is a pleasant surprise in a couple of ways. First, edited volumes often end up being mélanges of contributions united by little more than their bindings. Articles in this volume, however, cover the topic well and in a logical order. In a refreshingly concise introduction, the editors lay out the rationale for the book, its organization, and the main points of the chapters and how they relate to one another. Second, the book is worthwhile. As one of the semi-unemployed "soldiers" in the struggle that led to transport deregulation, I appreciate efforts to breathe life into the subject, although I begin to question the relevance of an event older than most of my students and some of my clothes. But the volume analyzes a range of issues that, in many cases, have continuing importance within transportation. Of perhaps even more significance, the volume is a rich case study of the impacts of regulatory changes on labor and management compensation and employment levels, as well as working conditions. As such, it would be of value to researchers and practitioners in transportation, human resource management, safety, and industrial organization.
The book may be thought of as being in three sections: 1) safety; 2) employment, productivity, and working conditions; and 3) compensation. All three sections have merit. If they had to be ranked, safety is the strongest and compensation the weakest.
Safety: Chapters 2 and 3
In a 23-page tour de force, Ian Savage examines trends in injuries since deregulation in trucking, railroads, and airlines. His command of the issues and data is impressive. Of particular importance, he makes the almost always neglected point that crash data alone provide an incomplete picture of worker safety. In trucking and railroads, transportation accidents account for only 12% of lost workdays and a mere 4% of those in the airline industry. Safety trends are examined using a variety of measures, such as lost workdays per full-time employee and per unit of output. Comparisons are provided with either all private industry or manufacturing.
In the next chapter, Daniel Rodriguez et al. explore relationships between, on the one hand, motor carrier financial performance, firm size and type of operations, and driver payment method with, on the other hand, driver safety. Not surprisingly, the sometimes confounding effects of these relationships, as well as data limitations, sap the strength of the results. But the authors do reach some important conclusions and the study is of value as an example of a very competent empirical investigation of complex issues.
Employment, Productivity, and Working Conditions: Chapters 4, 5, and 6
Kristen Monaco and Dale Belman examine the impacts of technology on working conditions for truck drivers. The study presents considerable information about the types of technologies in use, who uses them, and to what extent. Most interesting are their findings with regard to the ability of satellite-based systems to substitute for driver experience and to increase revenue-miles.
Next, Nancy Johnson and Jonathan Anderson explore employment, productivity, and working conditions in airlines following deregulation. The piece is a rich source for data about airline profitability, bankruptcies, and output. Some of this could have been relegated to appendices, though all of us enjoy checking out when our favorite or most hated carrier went belly up. The authors clearly lay out how swings in employment and productivity relate to technological innovations, such as the hub-and-spoke system, and the health of the overall economy.
The analysis of changes in working conditions, though, seems to be trying a bit too hard to make a case for deterioration. For example, Johnson and Anderson show that between 1970 and 2001, average weekly hours worked by pilots increased from 29 to 42. While acknowledging that airline accident rates steadily declined over this period and that 42 hours may not seem a very high number, they point out that pilots may suffer from jet lag and that even slight fatigue may result in catastrophic miscalculations. They do not mention, however, that aircraft in 2001 automatically took care of many more pilot tasks than was the case in 1970. Suggestive of increasing strain on airline attendants, the authors point out that between 1995 and 2000 incidences of crews having to enforce discipline on unruly passengers rose from 140 to 266. They fail, however, to put these statistics into any context, that in 1995 there were 0.00002 such incidents per plane departure, versus 0.00003 in 2000.
Daniel Rich's investigation of productivity, technology changes, and labor relations is the best of this section. Incorporating rail, air, truck, and water transport, Rich explores the interplay of labor-saving and labor-using technologies on the sources of productivity changes and employment and compensation levels. The discussions of the setting, relevant theories, data, empirical approaches, and results are all first rate. It is simply an excellent treatment of an extremely worthwhile, but complex, topic. I recommend it to you and will insist on it for my students.
Compensation: Chapters 7, 8, and 9
Though not without considerable merit, for two reasons the weakest contribution in the volume is Stephen Burks et al.'s study of executive earnings in trucking in the era of post-deregulation. The first is not the authors' fault. Data problems limited the investigation to 1977 to 1986, too short a period to see the full effects, if any, of deregulation.
The second reason is another matter. After a brief dip past 1980, the authors find that executive compensation recovered and, thereafter, increased continually, despite much less rosy compensation trends for their subordinates. In their conclusions they state: "From 1985 onwards they [i.e., trucking executives] received pay increases in line with the wider boom in executive pay." I believe there is a simple explanation that did not enter into the empirical investigation nor discussion given by the authors. Joe, out on the loading dock or driving a truck, may have few alternatives outside the industry and his salary is pegged to variations in the fortunes of his company. But many of the skills needed for managing a firm are not industry-specific and, as such, executive compensation in trucking would be influenced by compensation levels for executives throughout the economy. The immediate post-deregulation dip in compensation might have reflected a transition from executives with skills of specific value under regulation to those better suited to lead firms in unregulated environments.
John Bitzan presents an analysis of compensation levels for low- and mid-level managers in airlines, trucking, and rail. Similar to Burks et al., Bitzan does not account for earning levels outside of a manager's industry when examining the impacts of deregulation on their earnings and, like Burks et al., he finds little or no effect. As if to atone for these "sins," the bulk of the chapter addresses whether the qualifications of managers in transport industries changed after deregulation, how these changes compared with those in nontransportation industries, and how managers were compensated for their qualifications. These are important questions. Bitzan deals with them well and gets some intriguing results.
The editors, John Peoples and Wayne Talley, finish off the batting order with an examination of motor carrier owner-operators serving port cities. It is a perfect finale for the volume. They examine how changes in regulations affecting maritime shipping and advances in containerization impacted one segment of the motor carrier industry. The study is competently done and its results of interest. Of perhaps more significance, their work challenges readers to consider the broader impacts of policy changes in one industry throughout the economy. This may be a self-serving attempt to set the stage for their next volume. At least that is something to be wished for.
Reviewer address: Richard Beilock, University of Florida, PO Box 110240, Gainesville, FL 32611-0240, USA. E-mail: email@example.com.
Statistical and Econometric Methods for Transportation Data Analysis
Simon P. Washington, Matthew G. Karlaftis, and Fred L. Mannering
Chapman & Hall/CRC Press
April 2003, 425 pages
Transportation analysis has changed dramatically in the last 50 years. For example, the sequential four-step model of travel forecasting is slowly being replaced by activity-based analysis in addition to changes in the types of techniques and methods used. We have seen cross-classification analysis replaced by linear regression, and statistical and econometric methods such as hazard-based duration models and structural equations have become an integral part of the methodological framework of travel forecasting. With this evolution in methods comes a tremendous need for books that synthesize and extend the usual statistical theory presentation to one suitable for application-oriented audiences. Washington et al.'s book provides an excellent and needed addition to this genre of texts.
The book catalogs many of the major modeling techniques used in practice, most of which are also an important springboard for more advanced theoretical and largely academic transportation modeling. In this sense, the book is an excellent addition to a practicing transportation analyst's library as well as a perfect companion to a first year graduate modeling or methods course including, for example, travel forecasting, safety, and traffic engineering. One of the book's most useful features is its singular focus on transportation. All of the examples relate to, and are focused on, transportation problems. As an added benefit, the datasets used to develop the examples can be accessed via the publisher's website (http://www.crcpress.com/e_products/downloads/).
Washington et al.'s book is organized into three major sections: the fundamentals, continuous dependent variable models, and count and discrete dependent variable models. The fundamentals section presents basic statistical theory and includes topics such as central tendency, variability, hypotheses testing, and nonparametric tests. Part 2 includes discussion of regression, simultaneous equations, latent variables, and duration models, as well as panel data and time series analysis. In Part 3, count data models and the now familiar discrete choice and discrete/continuous models are presented.
Deciding what to include in and what to leave out of a methods book aimed at a transportation audience is often very difficult. Practitioners may not have the appropriate statistical background to immediately grasp the main concepts without some review of basic theory, yet too much foundation material overlaps with many statistical texts already available to graduate students. In Part 1, much of the material covered is readily available in most introductory statistical texts. It is useful material even for some first year graduate students; however, I would have liked to have seen this material divided into two sections, with most of the basic material going into an appendix. Part 1 could then be expanded to include many of the new modern graphical display techniques (e.g., Wilkenson 1999; Heiberger and Holland 2004) and perhaps some discussion about software and statistical computing (e.g., Gentle 2002). Despite this relatively minor caveat, the material presented in Part 1 is well done, with examples that clearly link theory to practice.
Part 2 is where the book really begins to distinguish itself. The first third deals with the basic linear regression model and includes a fairly comprehensive presentation of regression theory, assumptions, departures from assumptions, and the practical aspects of manipulating variables and estimating elasticities. The remaining two-thirds of Part 2 is devoted to fairly contemporary modeling techniques, at least in terms of transportation practice.
Of the various chapters, those dealing with time series and panel data analysis are perhaps the weakest. The remaining chapters in Part 2 are well written and the authors have done an excellent job summarizing the major modeling approaches and the main assumptions for each of the modeling techniques. For example, in their chapter on duration models, they begin with a brief discussion of the Kaplan-Meier method (the predominant nonparametric model used in survival analysis); the authors then turn to a longer exposition on semi-parametric and fully parametric models. Each section begins with a presentation of the basic model, followed by an example, which helps to motivate the method's application. In the chapter on duration models, all of the modeling approaches are tied together with a brief discussion comparing the different techniques. Finally, the chapter ends with a discussion of the modeling assumptions, which very cleverly motivates several more complicated modeling approaches addressing, in part, violations of the basic modeling assumptions.
In Part 3, the authors limit their coverage to three very important modeling approaches. The first approach is used for response variables that are considered count data and include the family of Poisson and negative binomial models. The second approach focuses on discrete choice models and the final section on discrete/continuous models. All of these modeling approaches are accessible to practitioners and increasingly form the foundation for handling many types of transportation problems. As with Part 2, each chapter begins with a presentation of the model structure and concludes with an example highly relevant to transportation planning and engineering practice.
Overall, this text adroitly fills a very important niche between practice and theory. Although I would liked to have seen a few additional topicsfor example, more on contemporary graphical techniques and some elaboration on simulation methods, which are playing an increasingly important role in travel forecastingin general, the book is very well written. I recommend it for most transportation analysts and believe it to be a good, solid addition to the libraries of transportation graduate students.
Gentle, J. 2002. Elements of Computational Statistics: Springer Texts in Statistics and Computing. New York, NY: Springer.
Heiberger, R.M. and B. Holland. 2004. Statistical Analysis and Data Display, An Intermediate Course with Examples in S-PLUS, R, and SAS: Springer Texts in Statistics. New York, NY: Springer.
Wilkenson, L. 1999. Statistics and Computing, The Grammar of Graphics: Springer Texts in Statistics. New York, NY: Springer.
Reviewer address: Debbie A. Niemeier, Professor, Department of Civil and Environmental Engineering, One Shields Avenue, Davis, CA 95616. E-mail: firstname.lastname@example.org.
Transportation After Deregulation
B. Starr McMullen, editor
2001, 140 pages
$97.95 € 97.95 £64.95
The U.S. Railroad Revitalization and Regulatory Act of 1976 funded the reorganized bankrupt northeast and midwest railroads that formed Conrail. After the act went into effect, subsequent legislation initiated deregulation across the transportation industry and lifted most of the remaining motor carrier restrictions, including those imposed by the states. The six papers in this volume in the series, Research in Transportation Economics, all deal with the theme of transportation deregulation and regulatory reform and can be classified under one of the following topics: timing and impacts of deregulation, technological change issues, safety, and railroad mergers.
In the paper by Wesley Wilson and William Wilson, the authors discuss the impact on the marketing patterns of railroads of the Staggers Rail Act, legislation that deregulated the railroad industry. Implementation of the act resulted in lower rates for shipment of goods, especially grain products. The authors present an econometric analysis of rail rates between 1972 and 1995 that focuses on five grain commodities. Their empirical model includes demand, cost, and price relationships based on the New Empirical Industrial Organization (NEIO) models and a specification for regulatory regimes that uses a dummy intercept and a time trend for regulatory reform. Their results indicate the prices of the five commodities decreased over time, but the magnitude across these commodities differed, ranging from 40% to 71%.
Lawrence Wong's study assesses the effects of deregulation in the motor carrier sector by sorting out the independent impact of deregulation from those changes caused by the interaction of deregulation and new technologies. To do this, Wong used a translog cost function model with a time trend incorporated as a third-order truncated Taylor series expansion for data from 1976 through 1987. Additionally, the modeling efforts allowed for the decomposition of the technological change into three components: input bias, output bias, and characteristic bias. The author attempts to test the Schumpeter hypothesis for the less-than-truckload (LTL) sector, but the evidence to support this hypothesis in the LTL sector was not present. The empirical results show that although implementation of new technologies provided labor-saving advantages, they required greater expenditures and induced input biases because of shifts in the output level. The results also showed that the LTL sector of the motor carrier industry is capital-intensive, which creates higher barriers for entry.
The paper by Kristen Monaco and Taggert Brooks presents a unique analysis of how deregulation affected wages in the motor carrier sector. The authors used a time series approach because prior transportation wage studies that employed cross-sectional methods could not control for macroeconomic effects. Motor carrier wages are modeled as a function of manufacturing wages, and the relationships between the two series are then assessed. Their results show that the effects of deregulation on wages in this sector was felt most strongly between 1980 and 1984 and that wages in this sector declined from 1972 to 1996.
Atreya Chakraborty and Mark Kazarosian do not explicitly address the productivity in the motor carrier industry in their paper, but instead, their analysis assesses the relationship between marketing strategy and information technology (IT). Some firms are likely to use IT to increase the timely delivery of goods, while those that may produce the same level of output (measured in ton-miles) as their IT-using counterparts may be less concerned with timeliness than they are with lower rates. Given these two approaches, which are aggregated into one dataset, analysis of motor carrier productivity is inconclusive without controlling for marketing strategies.
Although the financial condition of railroads has improved significantly since 1980, policymakers are rightly concerned about the future vitality of the sector if costs are not contained. C. Gregory Bereskin's paper focuses on the potential for transcontinental railroad mergers, a controversial issue given the small number of railroads in service and the concern that mergers could result in a monopoly. Mergers would likely increase efficiency and provide a higher quality of service, thus attracting more shippers and raising rail revenues. More specifically, the empirical analysis reveals that unexploited economies of scale could result in monopolistic pricing, a standard theory in market structure research.
In the final paper in this volume, Frank Rusco and W. David Walls examine the effects of deregulation on safety operations. Opponents of freight deregulation often assert that it would create increased competition but at the cost of making freight transportation less safe. The authors apply this hypothesis to the minibus market in Hong Kong, which is both regulated and unregulated. The authors' model shows that minibus drivers in the unregulated market tend to drive faster and experience higher accident rates than their regulated counterparts.
Deregulation in the transportation sector of the United States has resulted in great savings and increased flexibility. For example, freight transportation costs dropped sharply. Railroad rates fell from 4.2¢ per ton-mile in the 1970s to 2.6¢ per ton-mile in 1998. In addition, the railroad industry became more profitable. The cost of shipping by truck fell by $40 billion from 1980 to 1988, and deregulation has improved flexibility and enabled businesses to provide timelier deliveries, which contributes to a reduction in inventory costs. Despite the multitude of positive benefits from deregulation, there are some interesting policy issues that resulted, and these papers provide further empirical assessment. While readers may choose to peruse papers of topical interest, it would be valuable to read all of the papers because they provide current research on the effects of deregulation on the industry.
Reviewer address: Brian Sloboda, Bureau of Transportation Statistics, U.S. Department of Transportation, 400 Seventh St, SW, Room 3430, Washington, DC 20590. E-mail: email@example.com.