3. Theoretical and Freight Transportation Studies

The previous section focuses on literature that assesses the direct and indirect economic impacts of transportation investment, particularly investment in highways. Although methodologies that have been used to quantify the productivity effects of investments in infrastructure provide a wealth of information on the relationships between transportation and industry structure and performance, most of the work reviewed in Section 2 says little about the specific role of freight transportation. Likewise, the methodologies that have been used to quantify productivity impacts are not based on benefit-cost analysis. Consequently, to better inform this Freight BCA Study, literature specific to freight transportation (such as that on logistics management) and literature on economic impact analysis methodology are reviewed and evaluated in this Section of the Compilation Report.

3.1 Introduction

A number of themes emerge from the literature that focuses on 1) theoretical approaches to assessing the economic impacts of transportation and 2) the associated effects or issues attributable to freight transportation.

• First, there is a body of work on broad policy issues, trends and barriers to economic development. More recent papers discuss the evolving transportation sector driven by e-commerce and globalization.
• Second, a set of reports discuss methodologies and approaches for structuring and conducting benefit cost analysis, including the use of multi-criteria approaches, quantifying all types of external costs to reflect full marginal costs. On a related theme, works focusing on economic productivity and development quantify in various ways the effects (including empirical estimation) of public investment on economic performance and productivity.
• Third, a body of work addresses the impact and response of logistics systems to highway improvements, including inventory, warehousing, and supply chain management.
• Finally, a group of papers deals specifically with advances in technology, including benefits of ITS, policy, standards, and interoperability of trucks and inter-modal facilities.

3.2 Theme 1: Policy and Trends in Freight Transportation

Several papers focused on policy issues, trends and barriers to economic development, and the evolving transportation sector driven by e-commerce and globalization. Regan et al. (2000) discusses freight transportation and logistics trends and challenges in the new millennium. The paper identifies the key issues that affect freight planning and logistics. These include: increased demands for freight transportation and logistics services, and the ability of the physical and information infrastructure to meet these demands; the role of road pricing in urban freight transportation; the impact of information technology on goods movement; and new development in logistics management. Their work draws significantly on Delaney's 10th Annual State of Logistics Report "A Look back in Anger at Logistics Productivity". Delaney also recently delivered the 11th Annual State of Logistics Report, where he discusses logistics trends and the impact of the Internet. There he reports that business logistics costs were approximately 10% of GDP, with transportation costs continuing at a nominal 6% of GDP for the 7th year in a row. Third party logistics firms are experiencing significant growth in the US.

Specific freight industry attitudes towards policies to reduce congestion are reported by Golob (2000) in an extensive survey of California-based carriers. Given that freight transportation plays a vital role in the economy of the nation and of the state of California in particular, the 1998 California Transportation Plan for goods movement developed by the California Department of Transportation (Caltrans) identifies four constraints and deficiencies affecting freight transportation in the state: 1) capacity and congestion, 2) safety, 3) geometry and surface conditions, and 4) intermodal connections. This study addresses the first and last of these factors and touches on the second and third, from the point of view of trucking companies. It examines the impact of congestion on trucking operations, the use and usefulness of information technologies in their operations, and the value and efficiency of intermodal transfer facilities across the state. Results could help structure case studies.

In European Transport and Communication Networks, and Transport in a Unified Europe Banister et al. discuss evolving European/EC networks together with policy responses to a very broad range of issues. The collection of papers covers trans-European networks, the effects of dissolution of borders and remaining barriers, private sector investment and network diversity. They conclude that there needs to be a European strategy for transport and communications infrastructure investment, agreed by all member states and neighboring countries. They suggest moving away from increasing the physical capacity and extent of the network to a broader range of options including means to limit growth in demand through pricing, regulation and management, to optimize freight fleets and multi-modal transport, and to explore private sectors' role in new developments.

In sum, works on policy and trends provide the context for benefit cost analysis of freight transportation improvements. The main message is that cooperation between private and public sectors will be needed to ensure a transportation system that meets the evolving freight needs of business and consumers. With increasing demand for transportation, it is ever more important to have a sound framework for evaluating all costs and benefits.

3.3 Theme 2: Theoretical Approaches to Benefits and Productivity Assessments

Within this theme, the literature focuses on macro-economic, micro-economic, and overall cost-benefit analyses, as well as estimation of parameters.

3.3.1 Macro-Economic Analyses

Studies by and for the Federal Highway Administration Office of Policy Development have documented the effects of public highway capital on logistics system and commercial sector economic performance. In particular, Jacoby notes that Nadiri's 1996 study provides empirical evidence of the contribution of highway capital on the total output growth and productivity of 35 industry sectors of the US economy. This work and others on productivity studies are reviewed in more detail in Section 2 of this report.

From the perspective of methodologies employed to assess the productivity impacts of transportation investments various approaches have been employed by researchers. For example, Bell (1994) reviews macroeconomic analyses of the linkages between transportation investments and economic performance. Xin (1996) uses an input/output model to study regional economic benefits of transportation system projects. Duffy-Deno models the relationship between capital stock and per capita income as an economic development indicator.

A recent comprehensive review by Khanam examined empirical work on the relationship between highway capital stock and the output and productivity of goods-producing industries. In the published literature, the bulk of which is based on US data, the impact of public capital stock of various types on the output and productivity of different economic sectors has been examined. As discussed in Section 2, the evidence from these studies suggests that a positive relationship exists between public highway capital and private sector output and productivity; and the estimated size and significance of this relationship are very diverse and depend to a large extent on the approach followed. The results, expressed as output elasticities, range from 0.04 to 0.56; in some models, the estimates are statistically insignificant (from zero) or negative. The report compares results (output elasticities) obtained using Cobb-Douglas and Translog models.

Keeler (88) uses a translog cost function econometric model for an analysis of the benefits of Federal-aid highway infrastructure investments in the United States on costs and productivity of firms in road freight transport industry. The average sum of marginal benefits across all industries is about 0.294. This means that a $1 increase in net capital stock generates approximately $0.3 of cost saving producer benefits per year. These benefits continue over the design life of the road improvement.[14]

3.3.2 Micro-Economic Analyses

Of the papers reviewed in this report, Mohring and Williamson (1969) is the first formal analysis of what has been termed "reorganization" benefits. These benefits result from adjustments in logistical arrangements that shippers make in response to lower costs of freight movement. Typically, such adjustments would involve fewer warehouses and more miles of truck movement as shippers take advantage of lower freight costs to consolidate storage facilities and reduce inventory costs. These effects are the principal source of benefits not captured in the conventional approach to benefit-cost analysis.

The main conclusion of this paper is that the commonly used consumer's surplus measure of benefits is just as valid for transportation investments (including possible logistics re-organization) as it is for other types of investments. Regardless of whether situations reflect fixed or varying output level, derived demand schedules for transport inputs provide an exact measure of benefit from transport improvements. In fact, Mohring generalizes this result by going one step beyond to include cost reductions of inputs that were formerly not employed in the production process.[15] Using a simple production model, Mohring derives a quotient relationship between direct and total benefits for different scale economies. This theorem demonstrates that the magnitude of indirect benefits can be of the order of 12 percent of direct benefits.

Mohring's paper provides the theoretical foundation for the cost benefits analysis micro-economic framework. It demonstrates the validity of using consumer surplus in estimating net benefits of transportation investments under very broad conditions.

Subsequent work by HLB on behalf of the ATA Foundation addressed, in a micro-economic framework, the policy impact of full social cost pricing on the trucking industry. In response to the 1997 Highway Cost Allocation Study, this work considered both the positive and negative externalities of trucking, and found that the policy risk associated with social cost pricing goes beyond the transport sector and could affect productivity and competitiveness. The analysis is based on a partial equilibrium model of demand for trucking services and quantifies the change in consumer surplus as a result of tax collections. The model was implemented within a risk analysis framework accounting for uncertainties in key assumptions.

3.3.3 Benefit/Cost Analysis

Forkenbrock (1990) discusses two types of benefits of corridor highway investments, reductions in transportation costs and increases in economic activity. Although road user benefits should form the basis for decisions of whether or not to invest public funds to upgrade highways, policy makers often wish to see estimates of economic development. The question of which benefits should be taken into account depends on the geographic perspective, whether regional (where transfers can occur), or national (where road user net benefits can be demonstrated). Economic development impacts to corridors were estimated using IMPLAN, an input/output model. This paper is consistent with the micro-economic framework. Regional economic impacts, if reported at all, should be seen as adjunct information. To include them depends primarily on the geographical perspective.

Although the existence of positive network externalities remains a matter of debate in transportation circles, their quantitative significance is not doubted in related fields, such as telecommunications. Capello and Rietveld make a compelling argument for government policy as a means of correcting for an under-supply of highway infrastructure due to the existence of positive network externalities. By way of example, the authors argue that logistics-oriented telecommunications systems are characterized by positive externalities in the adoption process and, given the high fixed costs of acquisition, government policy might be justified in order to ensure the economically optimal critical mass of users. Other examples point to positive externalities in vertically integrated sectors where improvements for shippers in forward markets generate un-priced advantages for shippers in backward-linked markets. Government policy might be able to exploit such positive spillovers with policies that accelerate the take-up of advanced logistics in forward markets. HLB (for the American Trucking Foundation) found that even small positive network externalities per truck-mile could add up to an aggregate sum as large as the negative congestion externalities attributed to heavy trucks.

There are only a few studies that contain the basic theoretical and conceptual discussions underlying the approach proposed for FHWA's Freight BCA Study to evaluating benefits of freight-transportation improvements. The amount of key research is small because the great preponderance of the work on highway benefit-cost analysis simply does not address the issue of improvements for freight carriage. And, even where benefits to freight carriage are addressed, the treatment is usually incomplete, since effects of shippers' longer-run responses to lower freight costs are not, in most of the literature, considered.

3.3.4 Parameter Estimation

A number of studies focus on empirical estimation of elasticities. Bjorner makes an explicit distinction between traffic (km) and transport demand (ton-km), where the former is an input in the shipper's production of transport services, while the latter is derived from firm's production of output. Estimated price elasticity with respect to traffic is shown to be considerably higher than elasticity with respect to transport.

The most extensive review of alternative demand models and their elasticity estimates is given by Oum (1989) and Oum et al. (1992). Transport demand models of four types are described and elasticities for all commodities are provided for rail and truck transport with respect to freight rates, speed, and reliability. Oum distinguishes between elasticity of aggregate market demand, mode specific elasticities, and mode choice elasticities. Goodwin examines elasticities of transport demand with respect to petrol price and transport price for bus, rail and metro.

Rolle (1997) develops a model for the computation of demand for rail passenger travel as a function of own price and other mode prices (car, bus, plane etc) for the Swiss railway. Additional factors included in the model are rail and road network density, density of population and other socio-economic variables. This work is relevant only in that it provides an econometric approach to elasticity estimation. Roson (1995) models transportation cost as a concave function of distance and determines a method to find profit maximizing fares. The article is informative more than directly relevant.

The need for freight transportation data as well as current sources and challenges are summarized in a TRB report by Hancock. Policy issues for improved data management and sharing are discussed and a list of data sources is provided.

Elasticity estimation approaches need to be reviewed as to the most applicable and robust method for the BCA study. In the case of quantifying the re-organization effect, elasticity estimation was to be based on survey interviews and case studies. These need to be compared to baseline estimates obtained by other means.

3.4 Theme 3: Logistics

A number of studies focus on quantifying the relationship between transportation and other logistics processes. This includes:

• the interaction of transportation and inventory decisions (Constable),
• estimating the effects of carrier transit time on logistics cost and service, and
• the scale of indirect logistics savings as a function transport cost reductions.

Blanchard et al. explore relationships between highways and production costs through case studies, tactical logistics models, and theoretical treatments. The most compelling and general work is Mohring. (Case Studies) He also summarizes findings from Case studies carried out by KPMG for Transport Canada who have investigated, in a qualitative manner, the relationship between highways and logistics performance.

McKinnon (1996) examines the relationship between traffic growth and production and changes in logistics practices in the UK. Traffic growth for trucking is shown to be the result of logistics management interactions between planning, supplier/distributor choices, production scheduling, and transport fleet management. He shows that traffic growth is not simply a function of GDP growth, but also of restructuring effects and just-in-time freight scheduling. This paper might provide useful background information to help structure case studies.

3.5 Theme 4: Advances in Technology and Intermodal Freight Transport

A body of work deals with advances in technology, including benefits of ITS, policy, standards, and interoperability of trucks and inter-modal facilities. While conventional efforts to improve trucking productivity have typically focused on the separate components of the system, Fawaz explores combinations of trucks, roads, and operations at the system level to achieve productivity gains. An evaluation framework and model is demonstrated through a specific case study.

Various papers explore the impact of new technologies. The 1990 TRB report on new trucks for greater productivity explores the Turner proposal for new truck configurations. Effects related to safety, road wear, and overall operating costs and required infrastructure investments are quantified as compared to present standards.

An earlier 1980 USDOT study examined the impact of technological change on regional productivity. Although the study could only find a weak impact of technological change on a regional economy, it did highlight the importance of urban highways in the shaping of urban form. Another study examined the inter-operability of vehicles among NAFTA countries. Recommendations for making motor transport under NAFTA more efficient are addressed. Norris (1996) evaluates the status of intermodal freight in the US in general, while Zavattero discusses integrating intermodal freight needs into the transportation planning process. Muller (1997) provides an excellent overview of intermodal developments, movements, technologies, barriers, and future trends. It appears to be [the] reference on inter-modal transport operations.

Studies on advances in technology and intermodal developments provide useful background information, but do not contribute directly to the development of a comprehensive BCA framework. A number of benefits of intermodal freight capacity were identified by Norris including congestion reduction, lower emissions, higher fuel efficiency, greater safety, reduced highway deterioration, cost savings and greater systems efficiency. Congestion reduction is achieved by diverting freight traffic away from highways. For each 10 containers carried on intermodal rail, a minimum of 7 trucks are taken off highways. Lower emissions are a result of a five fold lower hydrocarbon emission per ton-miles for rail as compared to truck.

The main message of intermodal transport is that for long haul trips rail may be more efficient and cost effective provided intermodal capacity exists. Challenges for intermodal freight include regulatory barriers, operational, structural, and technological limitations, and demand driven impediments.

3.6 Relevance of this Body of Literature to the Freight BCA Study

The core literature on theoretical treatment of freight improvements, so far identified by this project team, and reviewed in the initial Microeconomic Framework White Paper consists of the following items:

• Herbert Mohring and Harold Williamson, Jr., "Scale Economies of Transport Improvements," Journal of Transport Economics and Policy, Volume 3, Number 3, September 1969.
• D. A. Quarmby, "Developments in the Retail Market and their Effect on Freight Distribution," Journal of Transport Economics and Policy, Volume 23, Number 1, January 1989.
• Herbert Mohring, Transportation Economics, 1977, especially chapters 8 and 11.
• Hickling Corporation (with Charles Rivers Associates and Christensen and Associates), Methodologies for Evaluating the Effects of Transportation Policies on the Economy, Technical Report, supplement to NCHRP Report 342, March 1991, especially Chapter 14 and Appendix D.
• Hickling Lewis Brod Inc., Measuring the Relationship between Freight Transportation and Industry Productivity (NCHRP 2-17(4)), June 1995.
• G. Blanchard, Highways and Logistics and Production Performance, Transport Canada/Economic Analysis Special Infrastructure Project, Report TP 12791E, June 1996.

All of these papers make an important contribution. Note, however, that the Mohring and Williamson paper is the first formal analysis of what Herbert Mohring refers to as "reorganization effects"—the adjustments in their logistical arrangements that shippers make in response to lower costs of freight movement. Typically, such adjustments would involve fewer warehouses and more miles of truck movement as shippers take advantage of lower freight costs to consolidate storage facilities and reduce inventory costs. These effects are the principal source of benefits not captured in the conventional approach to benefit-cost analysis.

The most complete, to date, theoretical and mathematical treatment of these effects is in the supplement to NCHRP 342, and particularly in Appendix D. Much of this material is highly technical and not easily accessible to a reader without a strong knowledge of micro-economic theory and a good grasp of mathematics. Quarmby's article describes reorganization effects in a manner readily comprehended by non-specialists. Quarmby wrote this paper, not as an economist (which he is), but as a senior executive of a major British supermarket chain considering the interplay of transportation and logistics costs. Parts of NCHRP 2-17(4) also provide a non-mathematical discussion of some of the issues.[16]

The main and most important result is the quantification of "reorganization" benefits. It was developed by Mohring and later reviewed by Blanchard. Based on Mohring's work, indirect savings as a fraction of direct savings are shown at Figure 1. As demonstrated by Mohring, these savings are a function of manufacturing scale economies. One very interesting insight is that, the greater the manufacturing scale economy, the lesser the relative importance of the indirect benefits. This result, which on the surface may be counter-intuitive, is explained by Blanchard as follows: Industries with large scale economies have a more transport intensive cost structure; before the cost reduction, they would already be operating from fewer plants, each with a large market area. As a consequence, there is relatively little room for restructuring their production and to accrue indirect savings. On the other hand, the relative importance of indirect benefits is greatest for industries with small-scale economies, a significant portion of the US economy.

Exhibit 1: Indirect savings as a fraction of direct savings for given transport cost reductions and manufacturing scale economies

The studies reviewed on policy and trends provide the context for benefit cost analysis of transportation improvements. With increasing demand for transportation, it is ever more important to have a sound framework for evaluating all costs and benefits and to be able to rank initiatives in terms of their value generation.

Research on theoretical approaches to cost benefit analysis (micro, and macro-economic) and logistics are complementary to what had already been examined as part of the white paper. The framework should be able to account for, at a high level, the impact of a wide number of initiatives, whether they are infrastructure related or of a technological nature.

A number of studies have been identified as relevant for the meta-analysis. Oum, and Oum et al. provide the most comprehensive review of both methods and actual elasticity estimates. For truck freight, demand elasticities are given by estimation method used and some by individual commodity. Having reviewed over 60 empirical studies of transport demand, Oum notes a wide range of values across different commodity groups, but also for the same group using different functional forms. The preliminary meta-analysis reveals that the BCA framework will need to work with elasticity ranges and uncertainty in the estimates.

Although not directly applicable here, a number of studies cover demand elasticities for other modes of transport. Goodwin examines elasticities of transport demand with respect to petrol price and transport price for bus, rail and metro using a micro-cost function; Rolle compares the elasticity of railway and car demand with respect to price. Using a bootstrap procedure, he obtains a distribution of these estimates. This procedure could be investigated for the present study.

Related data on external costs of inter-city trucking are given in Forkenbrock based on various sources. Blanchard reports illustrative logistics cost elasticities with respect to service levels, changes in reliability and lead time as part of his logistics analysis and review. Tyworth estimates the effects of carrier transit-time performance (mean delivery time and variation) on total logistics cost and service using an enhanced single-echelon continuous review inventory model. Finally, the so-called Mohring theorem allows the estimation of direct and indirect benefits for various scale economies as a function of transport cost reduction.

3.7 Concluding Remarks

Overall, this body of literature has confirmed the basis for the micro-economic framework had been identified and reviewed as part of the White Paper and initial developments. Complementing previous work, a number of papers have been found as key works for the meta-analysis, allowing the extraction and estimation of parameters for the BCA model. A variety of alternate approaches to elasticity estimation have also been found and will form the basis for the next phase of the project.

previous | next