Impacts of Policy-Induced Freight Modal Shifts (2019)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

33 Macro Mode Share Factors This section focuses on macro mode share factors, which are driven by major trends affect- ing mode share, such as productivity, deregulation and regulation, containerization, double stacking, just-in-time delivery (JIT), fuel costs and climate change, and international trade. Following this discussion of macro mode share factors, this report will focus on individual shipment mode choice factors, which are influenced by the specifics of individual commodity movements. These include commodity characteristics (shipment size, package characteristics, shipment shelf life, shipment value, and shipment density); shipper and receiver characteristics; access to modes; logistics costs (order and handling costs, transportation charges, capital carry- ing cost in transit, intangible service costs, inventory costs, loss and damage costs, and service reliability costs); and additional factors (length of haul, shipment frequency, and environmental/ sustainability). Productivity Productivity can be a key factor in mode choice. If productivity in one of several compet- ing freight modes increases relative to another mode, the more productive mode will have relatively lower costs and the ability to lower its prices to capture more traffic. Thus, differing productivity trends among competing modes can lead to mode shifts. In the last 25 years, labor productivity growth for railroads greatly surpassed labor productivity gains for other modes and in the overall economy. Moreover, multifactor productivity gains for rail, spurred by improve- ments in capital inputs and the organization of service delivery, far outstripped those in the private business sector (Apostolides 2003). Figure 24 provides indexes of labor productivity for a variety of transportation industries for which the Bureau of Labor Statistics publishes data, as well as for the business sector as a whole. Modes include air transportation (passenger and freight, combined), line-haul rail- roads, general freight trucking (long-distance), used household and office goods moving, postal service, and all businesses. Figure 24 also shows that rail productivity gains were substantial during the 1987 to 2004 period, but have since leveled off. Air transportation experienced marked gains during the 2000s, while trucking and the postal service have experienced gains that are lower than businesses as a whole. Household movers have seen substantial labor pro- ductivity decreases. According to the U.S. DOT, between 1987 and 2012, output-per-hour worked more than doubled in line-haul railroading. Line-haul railroads do not include switching and terminal operations or short-distance/local railroads. Long-distance, general freight trucking grew by C H A P T E R 3 Factors That Influence Freight Mode Choice

34 Impacts of Policy-Induced Freight Modal Shifts 39 percent over the same period. However, in recent years trucking has grown more rapidly. Long-distance, general freight trucking establishments exclude local trucking and truck opera- tors that require specialized equipment, such as flatbeds, tankers, or refrigerated trailers (Strocko et al. 2013). Figure 25 provides indexes of multifactor productivity for air transportation (passenger and freight combined), line-haul railroads, pipelines, and all businesses. As was the case for labor productivity, Figure 25 shows that rail has shown large increases in multifactor productivity, while multifactor productivity for the remaining available modes has increased at rates more in line with the economy as a whole. According to BTS (2010), increases in rail transportation multifactor productivity stem from technical progress, such as improved capital inputs and technological changes in the form of improved methods of service delivery. Improved technology for locomotives, freight cars, track, and structures has increased reliability and reduced maintenance needs. Improved information technology has also improved operational efficiency. Industry restructuring, including mergers, has permitted greater efficiency of labor and rail traffic moving over longer distances without interruption. Railroad company consolidation has led to more efficient use of equipment and lines (Apostolides 2003). The pipeline industry relies on a system that for the most part was installed during the middle of the last century. However, despite the system and capital infrastructure maturity, recent changes in the industry have allowed it to reduce labor and energy inputs to post large increases in productivity. One speculation is that increases in computerization have allowed reductions in personnel and increased energy efficiency, maximizing output in a mature industry (Jack Faucett Associates Inc. 2007b). Source: (Bureau of Labor Statistics 2014) 60 80 100 120 140 160 180 200 220 240 260 Air (LP) Rail (LP) Truck (LP) Moving (LP) USPS (LP) All Business (LP) 20 12 20 11 20 10 20 09 20 08 20 07 20 06 20 05 20 04 20 03 20 02 20 01 20 00 19 99 19 98 19 97 19 96 19 95 19 94 19 93 19 92 19 91 19 90 19 89 19 88 19 87 Figure 24. Labor productivity for transportation industries (1987–2012).

Factors That Influence Freight Mode Choice 35 Deregulation and Regulation Over the last quarter of the twentieth century, Congress has sharply curtailed regulation of transportation. Major legislation included the following: • Railroad Revitalization and Regulatory Reform Act of 1976 (the 4-R Act) • Motor Carrier Act of 1980 • Household Goods Act of 1980 • Staggers Rail Act of 1980 • Bus Regulatory Reform Act of 1982 • Surface Freight Forwarder Deregulation Act of 1986 • Negotiated Rates Act of 1993 • Trucking Industry Regulatory Reform Act of 1994 • Interstate Commerce Commission Termination Act of 1995 This legislation successively deregulated, either totally or in large part, trucking, railroads, bus service, and freight forwarders. In addition, this legislation lifted most of the remaining motor carrier restrictions, including those imposed by the states. Deregulation of motor carriers became complete, except for household movers, at the end of 1995. At that time, the federal govern- ment also preempted state regulation of trucking, eliminating the last controls over price and service in the motor carrier industry. It eliminated the need for motor carriers to file rates and authorized truckers to carry goods wherever they wanted to serve. The acts gave railroads more freedom to price, except when “captured shippers” could show that they faced a single carrier without significant alternatives (Moore 2014). According to most experts and observers, deregulation has worked well (Moore 2014). For example, Figure 26 provides data on the freight railroad industry since the Staggers Rail Act of Source: (Bureau of Labor Statistics 2014, Jack Faucett Associates Inc. 2007b) 90 100 110 120 130 140 150 160 170 Air (MFP) Rail (MFP) All Business (MFP) Pipeline (MFP) 20 12 20 11 20 10 20 09 20 08 20 07 20 06 20 05 20 04 20 03 20 02 20 01 20 00 19 99 19 98 19 97 19 96 19 95 19 94 19 93 19 92 19 91 19 90 19 89 19 88 19 87 Figure 25. Multifactor productivity for transportation industries (1987–2012).

36 Impacts of Policy-Induced Freight Modal Shifts 1980 deregulated the industry, showing the tremendous positive effects, with productivity and volumes soaring while rates decreased substantially. The Surface Transportation Board reports that railroad rates fell by 45 percent in inflation- adjusted dollars from 1984 to 1999. The demise of the Interstate Commerce Commission at the end of 1994 eliminated many of the statistics previously collected on the motor carrier industry. However, limited data show that revenue per ton-miles, adjusted for inflation, continued to decline, falling by 29 percent from 1990 to 1999. In addition, service to small communities improved, shippers’ complaints against truckers declined, unionization of drivers declined (reducing wage differentials between nonunionized drivers and the general labor force), and the number of new firms increased dramatically (Moore 2014). Regulation and deregulation can be important factors in mode shares. One of the four key findings from the U.S Department of Energy’s study of freight transportation share and a low- carbon future was that “Major mode shifts are unlikely without substantial changes in costs or strong regulatory measures” (U.S. Department of Energy 2013). Deregulation has clearly played an important role in mode share over the last 25 years. Recent regulatory actions, such as hours-of-service regulations in the trucking industry and positive train control in the rail industry, continue to play an important role. Future regulatory actions, particularly those related to climate change, truck size and weight, and autonomous vehicles could significantly shape mode shares. Containerization Intermodal freight transport is the use of two or more modes to move a shipment from origin to destination. The concept of logistically linking a freight movement with two or more trans- port modes is centuries old; however, containerization revolutionized the shipping business by making it easier for goods to be transferred from one mode to the other. There is a relationship between transport costs, distance, and modal choice. Rail/waterways are less expensive for longer distances, whereas trucks are cost-effective for shorter distances. However, intermodal transportation offers the opportunity to combine modes and find a less costly alternative. Rail intermodal, the transport of shipping containers and truck trailers on railroad flatcars, has grown tremendously over the past 25 years. The railroad industry reported an approximately “Rates” are revenue per ton-mile; “volume” is ton-miles. Source: AAR Source: (Association of American Railroads 2011) Figure 26. U.S. freight railroad performance since the Staggers Rail Act of 1980 (base year 1981 = 100).

Factors That Influence Freight Mode Choice 37 fivefold growth in trailer and container traffic on the railroads from 1965 to 1995 (Dewitt and Clinger 2014). U.S. rail intermodal volume was 3.1 million containers and trailers in 1980, 5.9 million in 1990, 9.1 million in 2000, and peaked at 12.3 million in 2006. Intermodal volume fell sharply during the 2007–2009 recession, but by 2013, it had rebounded to a record of nearly 13 million units (Railroads 2013, Association of American Railroads 2014). Figure 27 provides a summary of the growth of U.S. rail intermodal traffic from 1989 to 2013 in millions of contain- ers and trailers. Through September, 2013, intermodal accounted for 22.6 percent of revenue for major U.S. railroads, more than any other single commodity group. Containers have overtaken coal, which historically has been the largest single source of rail revenue (Railroads 2013). One of the four key findings from the U.S. Department of Energy’s study of freight transpor- tation share and a low-carbon future was that “Different freight modes offer different special- ized services, limiting opportunities for shifting freight from one mode to another”; a second was that “Truck-to-rail modal shifts have the greatest overall potential for energy reduction.” These key findings illustrate why the trend toward containerization is so important for mode shift (U.S. Department of Energy 2013). Double Stacking Related to containerization, yet a unique trend in itself, is the emergence of double stacking. In 1990, containers accounted for 44 percent of rail intermodal volume. By 2000, the share was 69 percent, and in 2012, it was a record 87 percent. Unlike trailers, containers can be double stacked, thereby sharply increasing productivity and helping to ensure that there is sufficient traffic density to keep rail intermodal cost competitive with all truck movements. Transfer- ring containers to and from ships and trucks is also easier, further enhancing productivity (Railroads 2013). This important shift in the composition of the North American intermodal rail fleet took place in the 1990s, with the move away from piggybacking, or Trailer on Flat Car (TOFC), toward Containers on Flat Car (COFC). Figure 28 illustrates this trend over the 20-year period from 1981, when the intermodal rail fleet had virtually no double stacking, through 2001, when 60 percent of the fleet was double-stack. Source: (Association of American Railroads 2014) Figure 27. U.S. rail intermodal traffic 1989–2013 (millions of containers and trailers).

38 Impacts of Policy-Induced Freight Modal Shifts As the intermodal market evolved, equipment availability and standardization were constant issues. Although the intermodal trailer still plays a role in the market, the primary intermodal equipment type is the domestic container, with trailer traffic continuing to be converted. Rail- roads have recognized the train efficiencies in double-stack operations and have configured networks and terminals to accommodate them. Double stacking of containers saves much more convoy space than the piggyback method, with the added advantage of not having to carry a trailer. Intermodal transportation has traditionally lagged behind the highway industry in equip- ment size and capacity, but the 53-foot container has become the common piece of equipment (Prince 2010). The development of long-distance corridors linking major port gateways such as Los Angeles/ Long Beach to inland destinations spurred the setting of double-stack-unit train services. How- ever, many rail routes were not compatible with double stacking because of the required height clearance for bridges and tunnels. Converting a rail line to double stacking can be a costly undertaking. U.S. railroads have made such investments on high-priority corridors, raising clearances along rail routes to accommodate the additional height required (Railroads 2013). TOFC services that used to dominate have become marginalized because of the more efficient use of rail assets permitted by double stacking and the commitment of trucking companies to integrate their drayage services with long-distance intermodal rail services. In addition, most well cars (stake cars) can accommodate 53-foot domestic containers, undermining the need for piggybacking. What railroads used to carry as TOFC, they now carry as COFC. JIT JIT manufacturing or delivery is a production model in which producers create items to meet demand rather than creating them in surplus, or in advance of need. The purpose of JIT produc- tion is to avoid the waste associated with overproduction, waiting and excess inventory. Henry Ford described the JIT concept in his 1923 book, My Life and Work (Ford and Crowther 1923): We have found in buying materials that it is not worthwhile to buy for other than immediate needs. We buy only enough to fit into the plan of production, taking into consideration the state of transportation at the time. If transportation were perfect and an even flow of materials could be assured, it would not be necessary to carry any stock whatsoever. The carloads of raw materials would arrive on schedule and in the planned order and amounts, and go from the railway cars into production. That would save a great deal of money, for it would give a very rapid turnover and thus decrease the amount of money tied up in materials. Source: (Prince 2001) Figure 28. Composition of intermodal fleet by type of unit (1981–2001).

Factors That Influence Freight Mode Choice 39 Toyota adopted JIT in the Toyota Production System in the 1970s as a means of eliminating waste. However, it was not at the Ford Motor Company that Toyota representatives saw the JIT model in action. When Toyota toured plants in the United States in 1956, Ford had not yet fully implemented the JIT model. It was at Piggly Wiggly, the first self-service grocery chain, that Toyota representatives saw JIT demonstrated (Whatis 2014). According to observers, such as the Center for Climate and Energy Solutions, “The rise in popularity of JIT inventory management made businesses more reliant on timely delivery, favoring trucks since they have access to an expansive network of roads and can move product quickly. As a result, trucks gained more and more market share throughout the 1990s and into the 2000s” (Center for Climate and Energy Solutions 2014). Fuel Costs and Climate Change Different modes use significantly different amounts of energy per unit of freight movement and, as a result, produce different levels of emissions, including GHGs. Figure 29 provides estimates of energy use in BTUs per ton-mile of freight. As Figure 29 shows, rail and water modes are four to five times more energy efficient than trucking. Railroads spend relatively less on fuel, reflecting the economies of scale and corresponding fuel savings that rail achieves by hauling very large volumes of freight over long distances. The higher price of trucking services and the lower price of rail services reflect the differences in fuel use (Brogan et al. 2013). Freight railroads are also increasing fuel efficiency faster than trucking companies are. Table 8 shows the average annual percent change in energy intensity for trucks and rail from 1970 to 2011 and from 2001 to 2011. Rail has achieved a significantly higher decrease in energy use than trucks, especially on a ton-mile basis. For trucking, fuel costs in early 2010 accounted for 31 percent of marginal operating costs per mile; driver labor costs accounted for 36 percent. These proportions have varied considerably in recent years, with fuel accounting for 38 percent of marginal operating costs in 2008 and 28 per- cent in 2009. The cost and volatility of fuel prices from 2000 to 2010 have been major factors pushing the motor carrier industry to search for more fuel-efficient engines and transmissions, Source: (Oak Ridge National Laboratory 2013) 0 200 400 600 800 1000 1200 1400 Truck Rail Water 1285 289 217 BT U p er to n- m ile Figure 29. Estimates of energy use in BTUs per ton-mile of freight by mode.

40 Impacts of Policy-Induced Freight Modal Shifts more aerodynamically clean truck shapes, and more efficient head-haul and backhaul routing and dispatching (Brogan et al. 2013). To make their operations more fuel efficient, railroads have been moving longer distances between interchanges, buying more fuel-efficient locomotives, using innovative equipment such as aluminum freight cars and lightweight double-stack container cars, and reducing loco- motive idling time (Apostolides 2003). Figure 30 shows the impact of increases in diesel fuel prices on line-haul costs. The fuel- efficient rail and water modes, especially container on barge (COB), suffer less from fuel price increases than trucking does. The gap between line-haul costs for the truck mode and the modes of rail and water widens as fuel prices increase, such that shippers will be able to realize signifi- cant savings by diverting to rail and water. GHGs directly relate to fuel consumption; moving freight by rail instead of truck lowers GHG emissions by 75 percent (Association of American Railroads 2014). Figure 31 shows the relatively low contribution of rail to U.S. GHG emissions in 2012. Year Heavy Single-Unit and Combination Trucks (BTU/vehicle-mile) Class I Freight Railroad (BTU/freight-car-mile) Class I Freight Railroad (BTU/ton- mile) 1970–2011 -0.30% -0.60% -2.00% 2001–2011 -0.60% -0.70% -1.50% Source: (Oak Ridge National Laboratory 2013) Table 8. Average annual percent change in energy intensity (1970–2011). Source: (Maritime Administration and U.S. DOT 2008) Figure 30. Comparison by mode of the impact of diesel fuel prices on line-haul costs (2008 $).

Factors That Influence Freight Mode Choice 41 Higher fuel prices, the threat of climate change, the push toward energy independence and energy security, and the greening of the supply chain will all create pressure to invest in policies that shift freight from trucks to rail, water, and pipelines. International Trade Alterations in trade patterns affect the choice of transportation modes used in this country. For the exports of several countries, such as Canada and Mexico, the United States is a leading destination. This is also the case, to a lesser extent, for imports. Much of the trade between the United States and Canada is due to the integration of North American automotive produc- tion. The remainder of the trade is either raw or semi-processed materials (e.g., lumber and petroleum products) from Canada or the exchange of other manufactured products facilitated by strong bilateral relations and the proximity of the two countries. A notable share of the trade with Mexico is exports of automotive products and electronics to Mexico for assembly in Maquiladora factories and then re-importation into the United States as finished products. U.S. trade with Asian Pacific countries has grown, although those countries’ imports from, and exports to, the United States represent a smaller share of the total trade of the United States in comparison to trade with Canada and Mexico. As U.S. exports to Asian countries grow, contain- erized cargo transiting West Coast ports is likely to increase, creating more demand for efficient intermodal services. The BTS has identified a number of trends that are affecting international trade and freight mode choice (U.S. DOT 2003, BTS 2010). These include the following: • U.S. and foreign direct investment. A major factor influencing U.S. trading relations is the outflow of U.S. direct investment abroad, and the inflow of foreign direct investment in the United States. These trends are important because investment by businesses could comple- ment the flow of merchandise trade and affect transportation services carriers, such as ship- ping lines and airlines. Growth in these investments also results in increased intra-industry and intra-firm trade, as multinational companies establish branch plants or supply chains. Source: (Association of American Railroads 2014) Figure 31. U.S. GHG emissions by economic sector and transportation mode in 2012.

42 Impacts of Policy-Induced Freight Modal Shifts • Changes in U.S. reliance on imported consumer products. Over the past few decades, as U.S. population and income has grown, merchandise trade and freight movements have risen greatly. Shifts in age and geographic distribution, immigration, and participation in the workforce have combined to affect consumer tastes for foreign products, thus increasing the demand for traded goods and transportation services. If the U.S. population continues to grow at past rates and some of the observed shifts in geographic concentrations persist, demand for transportation services will increase, affecting both local freight movements as well as longer distance flows. • Shift in U.S. GDP toward services. As the structural composition of U.S. GDP has shifted toward more services, the nation’s reliance on imports for manufactured goods has increased. It is possible that these changes could continue to influence the volume of U.S. international trade and affect goods movement within the United States for many years to come. • Integration of the world’s economy. Although societies have traded with each other for millennia, the pace and scale of the integration of the world’s economy during the past five decades may well be unparalleled in history. Today, the world economy continues to change in dramatic ways. Due in part to lower transportation costs, geographic distance no longer protects industries from international competition as it once may have. The global nature of much of manufacturing makes it difficult to determine whether a computer is “American,” a car “Japanese,” or a television “Mexican.” Many expect globalization to continue to shape world economic activities, influence the location of goods production and distribution, and ultimately affect the transportation of goods into and out of the United States. Even if growth in the volume of freight moved were to taper off, ongoing changes in business logistics, out- sourcing, and JIT inventory systems that characterize global production could increase the demand for more frequent and smaller shipments. • Transportation and telecommunications technologies. Transportation and telecommuni- cations technologies have contributed to the rapid growth of international trade, helping overcome the resistance of space and time. These technologies have allowed the unparalleled mobility of goods and people. Air and containerized cargo improved merchandise trade dra- matically through advances in both vehicles and infrastructure, increasing speed, reliability, and safety, while reducing transportation costs. For telecommunications, improvements in voice, text, and data technologies have allowed fundamental changes in services trade, includ- ing transportation services. Improvements in information technologies make it easier to seamlessly coordinate transportation operations across physically distributed transportation networks, facilitate intermodal and multimodal movements of international trade, enhance transportation solutions for freight shippers, and allow significant gains in transmitting pre- clearance cargo and crew information for security operations. • Reductions in trade barriers. U.S. international trade has expanded globally, in part because of substantial reductions in trade barriers resulting from changes in policy. Reductions in trade barriers included the formation of regional economic groupings such as North American Free Trade Agreement (NAFTA), the European Union, and the MERCOSUR56 (Mercado Común del Sur means Southern Common Market) in Latin America. As trade barriers declined, the relationships between national governments and businesses changed, creating economic con- ditions that enhanced access to global markets and resources. Significant changes that could affect the economic deregulation of international transportation services, multilateral Open Skies agreements, privatization of infrastructure, and general agreements among World Trade Organization member nations could further facilitate trade interactions. Because of these trends, international trade will continue its expected growth, and the demand for improved intermodal access to U.S. ports will rise, particularly at containerized ports in urban areas. Issues and concerns include the condition of local roads accessing ports, at-grade rail crossings, rail drayage time and costs, dredging and channel depths, and availability of truck-only lanes for access to ports.

Factors That Influence Freight Mode Choice 43 Insights from the IDIs For freight mode choice policy to be effective in inducing desired changes in behavior, policy- makers must have a clear idea about how the intended target(s) would react to the policy measure(s). Lack of knowledge about the behavioral response of the target groups could lead to negative unintended effects or ineffective policymaking. IDIs involve conducting intensive individual interviews that provide the interviewee an opportunity to discuss a specific topic overall, without being constrained by specific questions. IDIs also provide the interviewer the flexibility to change the questions based on the inputs received during the IDI. Using IDIs, the NCFRP Project 44 research team collected information on the firms’ overall supply chain process, the factors considered in mode choice decisions, and policies/changes that they require or desire in the future for an efficient modal split. The following sections provide (1) a descrip- tion of the methodology and approach to IDIs, (2) an overview of the firms interviewed and the supply chain process, (3) the factors that interviewees consider in mode choice decisions, (4) the suggestions given by the firms to move toward sustainable freight mode shifts, and (5) concluding remarks that summarize the IDIs overall. Approach to IDIs IDIs are a qualitative research technique used to conduct intensive individual interviews exploring the interviewee’s perspective on an idea, program, or situation. These interviews are very useful in obtaining detailed information about people’s behavior and offering a more complete picture of their perspectives (Boyce and Neale 2006). Moreover, IDIs are important research tools that enable extended discussions with key individuals, an opportunity to gather insight, probe for additional information, and dynamically change the direction of the discus- sion as demanded by the circumstances. Typically, IDIs are 1 hour in duration and provide an excellent way to discuss complex problems and gain insight from decision makers, industry leaders, and lead researchers who can provide critical input to transportation problems. Com- plemented with surveys and field trips, IDIs give deeper insights about agent behavior. To obtain useful results from IDIs, the questionnaire needs to be well prepared and tested beforehand to avoid any confusion (Holguín-Veras et al. 2013). A survey script should be pre- pared with a brief introduction specifying the purpose of the interview. It is also important to mention the confidentiality of the responses. After explaining the process, the script should contain the main relevant (general and specific) questions about the topic of interest. This ques- tionnaire should be a guide, not a straitjacket. The interview should gradually transform into a conversation where more questions and answers may arise. The script should have a note mentioning that if the interviewee has more questions or suggestions, he/she should contact the interviewer. Finally, to avoid misunderstandings, the text/document produced based on the IDI should be shared with the interviewee to confirm that the written conclusions/ideas are exactly what the interviewee wanted to convey. This is important for ensuring that the results represent a consensus of what was discussed and that no confidential information is released. IDIs can be an efficient means of data collection in the freight industry. These interviews should be focused on individuals familiar with the topic of interest, whether they be company specific or people who work within a certain freight sector (Holguín-Veras et al. 2013). Also, IDIs can shed light on important freight decisions. For instance, if the main factors determining the freight mode to transport certain commodities are sought, an interview with a knowledgeable supply chain manager in a shipper company might be all that is required to determine those fac- tors. The NCFRP Project 44 research team conducted research on the context of policy-induced freight modal shifts, or more specifically, sought understanding of the factors that determine mode choice. The team interviewed representatives of the freight industry (selected shippers,

44 Impacts of Policy-Induced Freight Modal Shifts carriers, and receivers that collectively cover a broad range of business conditions and geogra- phies) about the types of freight mode choice considered when shipping or receiving goods. The interviews highlighted the key factors taken into account when mode choice decisions are made. The research team (1) defined the interview goals and objectives, (2) designed the script for the IDIs, (3) identified the various stakeholder groups to be targeted during the interview process (e.g., manufacturers in various industry sectors, trucking companies, railroads, and receivers of goods), and (4) produced a draft list of individuals considered for IDIs. Having identified the individuals selected for IDIs, the team proceeded to schedule conference calls, in-person inter- views, and field trips to gather input on the firm, the supply chain, and the factors taken into account when making mode choice decisions; problems or challenges faced; and policies or improvements recommended. After analyzing all the IDI responses, the research team selected the responses of 11 IDIs for further analysis. Description of the 11 Firms Interviewed Table 9 shows a summary of the industry sectors to which the selected group of interviewees belong and an overview of their mode choices. Table 9 shows that truck is the predominant mode choice, as it is used by all shippers and receivers (nine of the eleven interviewees), followed by intermodal (seven companies) and rail-only (six companies). Intermodal is used by most shippers (three out of five) and receivers (three out of four). Inland waterways, as well as air and courier services are used on a limited basis. All shippers and one receiver indicated that they use rail. Two shippers indicated they use inland waterways for shipping/receiving bulk products. Air is rarely used by shippers, unless the cargo has to be delivered urgently. Since the mode choices depend on the nature of the business and the specifics of the corresponding supply chains, brief descriptions of the companies are also provided below. Receiver 1 (R1) is a large national retailer that imports about 25 percent of the company’s cargo and sources the remaining 75 percent domestically. The primary modes used for domestic shipments are truck and rail. The company utilizes their distribution centers (DCs) to con- solidate cargo to dispatch full truckloads (FTLs) or full-train loads in order to be cost efficient. The typical shipment to stores is a full 53-foot trailer. When possible, the company aims to do Firm Industry Sector (NAICS 2-digit) T ru ck R ai l In la nd w at er w ay s O ce an In te rm od al A ir C ou ri er Receiver 1 Receiver 2 Receiver 3 Receiver 4 Shipper 1 Heavy Manufacturing (33) Shipper 2 Agriculture, Forestry, Fishing, and Hunting (11) Shipper 3 Heavy Manufacturing (33) Shipper 4 Light Manufacturing (32) Shipper 5 Light Manufacturing (32) Carrier 1 Carrier 2 Retail (44) Transportation and Warehousing (48) Table 9. Current supply chain patterns of the 11 selected interviewees.

Factors That Influence Freight Mode Choice 45 backhauls by coordinating multivendor pickups on return trips back to the DCs. Although the company owns the majority of the trailers used for delivery, the company also hires inter- modal carriers to move cargo by train and truck. These carriers make the day-to-day mode choice decisions based on the goals and cost structure specified in the contract. Receiver 2 (R2) is a chain of department stores that receives cargo from both international and domestic sources. These inbound shipments are about evenly divided between intermodal (rail and truck) and FTL. The receiver operates regional and local DCs and e-commerce fulfill- ment centers. Inventory from the DCs and store locations is used to fulfill e-commerce orders, although, in very few cases, orders may be fulfilled directly from the vendors. These shipments are carried out by third-party de-consolidators, major intermodal providers, and third-party logistics (3PL) providers. Every year, the firm aims to use intermodal as much as possible, espe- cially for trips longer than 700 miles, by issuing an early proposal to 3PLs. The average drayage distance for intermodal deliveries is 100 miles. Receiver 3 (R3) runs a chain of food stores and uses the firm’s DCs to transport goods to store locations. The DCs receive cargo using their own truck fleet or vendors’ trucks. A DC typi- cally receives 60 to 70 deliveries per day, of which 40 percent are perishable goods that require temperature-controlled facilities. Aside from trucks, the company receives two to three car loads per week via a rail 3PL provider. Truck is the main mode for deliveries to the stores. A typical store used to receive 10 deliveries a week due to perishable goods, but this has now been reduced to five deliveries due to the use of new technologies. The backhaul department controls 50 percent of the total deliveries to the DCs, of which 30 percent is the “true” backhaul by company-owned trucks, while the remaining 20 percent is contingency deliveries mostly due to issues with vendors. This firm has a department to manage recycling, waste management, and returns. Receiver 4 (R4) is a leading retailer that sources cargo domestically and internationally. Inside the United States, the major shipping modes are truck and intermodal. This firm oper- ates three types of DCs: rapid DCs for domestic products; stocking DCs for imports and sea- sonal domestic supplies to stores; and bulk DCs to handle oversize/weight cargo. In total, all of their DCs in the United States receive about 12,000 shipments and send nearly 10,000 ship- ments a week. Imports are received by ocean using major ports in the United States (East and West), accounting for nearly 120,000 40-foot equivalent units a year. The company spends about 65 percent of transportation costs on trucking, the majority being FTL; 10 percent is spent on intermodal; 13 percent of the spending goes to waterways for imports; and the remainder is spent on express carriers. The majority of the cargo goes through the company’s DCs before being transported to stores. Vendors supply about 25 percent of the domestic products directly to stores. Shipper 1 (S1) is a large machine manufacturer that receives inputs mostly from international suppliers. Outbound cargo are the final products, which are mostly transported internationally; 20 percent are over-dimensional loads. Freight is transported by all modes and is typically large. In general, small-sized products come in less-than-truckload (LTL), and medium- to large-sized shipments are delivered by FTL. The inputs to the manufacturing units are transported by LTL on a JIT basis as the firm finds it to be more cost efficient due to high inventory costs. For expedited services, the company uses air transportation. Rail is used when the product is too large to be transported over road, and the shipments are sent directly to the port. The company uses ocean as the major mode of transport (for both inbound and outbound), utilizing major ports all over the world. Consolidation takes place at ports before loading to the containers/ project vessels. Shipper 2 (S2) is an international provider of products and services in multiple economic sectors. The largest portion of transportation expenditure is spent on trucking, followed by rail and then intermodal. The rest of the spending is on barge, routing, and terminalling (temporary

46 Impacts of Policy-Induced Freight Modal Shifts storage for cargo when switching from one mode/vehicle to another). The company owns reefer trucks for which the drivers are outsourced. Forty percent of the cargo by value is refrigerated cargo, transported only by trucks. The majority of the liquid products are shipped by rail. For exports, the firm uses ocean, particularly for bulk freight. Shipper 3 (S3) is a large manufacturer with customers across the globe. The raw material for this firm is sourced from various vendors across the United States. The two most signifi- cant challenges faced by this firm are ensuring that the vast quantity of input materials arrive on time at the manufacturing sites and delivering on time to the firm’s numerous customers, particularly those with JIT deliveries. Due to large-scale operations, the supply chain needed to undergo a drastic change to accommodate customers’ requirements. The inbound deliveries exceed 25 million tons a year; 60 percent comes by truck, 30 percent comes by rail, and 10 percent comes via barge/vessels. The outbound deliveries are less than 25 million tons per year; 53 per- cent is delivered by truck, 39 percent is delivered by rail, and 8 percent is delivered by barge/ vessels. In terms of number of vehicles per year, this firm sends out more than 500,000 trucks, 90,000 railcars, and 1,500 barges. The company handles 90 percent of the outbound deliveries, and 3PL providers transport the remaining 10 percent. Shipper 4 (S4) is a leading mid-scale manufacturer, whose outbound cargo exceeds 100,000 tons a year. The largest portion goes by truck (75 percent FTL, 7 percent LTL), followed by intermodal (9 percent). The remainder goes by rail, ocean, and air. For international shipments, the firm uses a handful of ports in the United States. All FTLs make multiple deliveries to different receiv- ers, while some of the LTL shipments occupy less than half of the truck’s capacity (based on the customer’s order). The truck deliveries to customers are done either at night or early in the morning. This firm has its own trucks running between manufacturing units and warehouses. Most of the other trucks are managed by 3PL providers. The inbound cargo is both domestic and international, for which the entire supply chain is handled by the suppliers. Shipper 5 (S5) is a major manufacturer of chemicals and agriculture products. The majority of the inbound cargo is domestic, with 90 percent (by weight) transported by rail. One to five rail deliveries are received per week, and shipment sizes vary from 160,000 to 180,000 lb. The outbound shipment sizes range from 2,000 to 180,000 lb and are transported by truck (40,000–45,000 lb), rail, or ocean. Ocean is the predominant mode for outbound cargo as the main customers are international. The firm finds it not feasible to use inland waterways, but uses major ocean ports in United States. Volatile and hazardous material cargo are transported by rail because rail provides safer service than trucks for the volume of cargo handled. Carrier 1 (C1) specializes in providing a wide spectrum of transportation services for inland waterways. The services vary from entire logistics including transportation and storage, towing, and renting out equipment. The carrier deals with unique types of cargo that are oversized or overweight and that cannot be shipped by either rail or truck. This firm operates for 7 to 8 months in a year. Carrier 2 (C2) is a 3PL that provides end-to-end service, transporting goods door to door. They provide value-added services to their customers such as warehousing, packaging, and grading. The smallest shipment the company handles is one FTL, and it sends 40 to 50 trucks to customers on a regular day and 130 to 150 trucks on a busy day. In a typical week, the firm moves more than 100 railcars containing approximately 3 to 3.5 truckloads of cargo per railcar. This firm ensures reliability by tracking each railcar throughout the trip using GPS devices. The drayage distance is typically less than 350 miles. The description of the companies interviewed and their supply chains helps to better under- stand the factors that are taken into consideration when making mode choice decisions. The factors that arose during the interviews are presented next.

Factors That Influence Freight Mode Choice 47 Factors Affecting Mode Choice The IDIs revealed that various factors influence mode choice, as summarized in Table 10. It is worth noting that C1 and C2 operate in specific mode markets (inland waterways and intermodal/rail, respectively), so the information they provided should be interpreted as the particular factors they believe their customers consider when selecting the freight mode to use. This section discusses the most significant factors. Cost The IDIs revealed that cost is one of the top two influencers of mode choice; it was identi- fied by eight of the eleven interviewees as a key factor. However, it is worth noting that cost is never the only determinant of mode choice; it is usually associated with other factors because companies need to maintain a certain quality of service in their operations. So, if the cheapest mode option will decrease product quality or other operating standards below the minimum required by the company, that mode will not be selected. The companies aim for FTL over LTL or full-train loads instead of a single or a few rail cars, when possible, to increase efficiency and reduce costs. R2 stated that cost is the most important factor and that, as a result, it aims to use intermodal as much as possible to decrease costs. Quality of Service The second of the top two factors is quality of service, which was listed as a significant factor by eight of the eleven interviewees. The term “quality of service” was used to refer to the wide range of service characteristics that influence quality of service, including (among others) reli- ability, level of service, consistency, dependability, proper handling of the cargo, and good customer service. The interviewees stated that the need for high quality of service leads them to prefer trucks and that a negative view of the quality of service provided by rail and intermodal ID Factors R ec ei ve r 1 R ec ei ve r 2 R ec ei ve r 3 R ec ei ve r 4 Sh ip pe r 1 Sh ip pe r 2 Sh ip pe r 3 Sh ip pe r 4 Sh ip pe r 5 C ar ri er 1 C ar ri er 2 T ot al 1 Cost 8 2 Quality of service 8 2a Reliability (consistent on-time deliveries) 5 2b Level of service (on-time deliveries + overall service) 3 3 Product type 6 4 Seasonal changes (weather or sales periods) 5 5 Drayage distance to/from intermodal/marine terminal 4 6 Shipment distance 4 7 Shipment size 4 8 Land/infrastructure for inland waterways 2 9 Transit time 2 10 Inventory space at shipper/receiver location 2 11 Impact of another agent on mode choice 2 12 Impact of delays 2 13 Ability to track shipments 1 14 Backhaul availability 1 15 Cargo damage 1 Table 10. Factors influencing mode choice.

48 Impacts of Policy-Induced Freight Modal Shifts deters them from increasing the use of these modes. R3 stated that the company uses trucks to get deliveries on time and with the desired quality of service standards, given that the com- pany’s sales depend highly on the timeliness of the shipment: “If rail does not meet our sales timing, we are going to end up losing a lot of product. If we get it all on Monday and the sale starts on Saturday, we are in trouble. With trucks, the deliveries happen in a timely manner.” C2 expressed that the impacts of delays are important for its customers. Since a railcar accom- modates nearly 3 to 3.5 truckloads, the delay of a shipment by rail will have a larger impact on the receivers than if a truck were delayed. R3 expressed a preference for using a rail 3PL over rail because, from the company’s experience, the 3PL provides a higher quality of service and other value-added services such as sorting, packaging, and storage. R4 mentioned that since it only maintains 2 days of inventory, a delay on a key shipment could significantly impact operations. This is a major issue for rail, which is perceived to be unreliable. With the increased need for JIT shipments due to insufficient inventory space at receiving locations, shippers need to fulfill frequent small orders, which leads to the increased use of faster, more reliable modes to meet customers’ needs. S5 values cost over the quality of service because on-time deliveries are not a major concern, except for at a few locations where inventory management is a problem due to unreliable delivery times from rail. A recurring theme in the IDIs was the use of cost and quality of service together to select the mode. This is shown in Table 10 with six of the interviewees selecting both factors and three of the six explicitly mentioning that they choose the most cost-effective method that will get the shipment to the destination on time: “When it comes to modal selection, generally we will choose the mode that will deliver the most cost-effective solution in getting it [there] when we want.” However, having on-time deliveries is a key factor for time-critical shipments. In these cases, they will opt for more expensive options to ensure a timely arrival. S1 uses air when expedited shipments are required. Conversely, S4 said that if there is flexibility in delivery times, the firm opts for intermodal as much as possible. Product Type The type of product being shipped was mentioned by six of the interviewees as a key factor in mode choice. R1 indicated that specialized products (e.g., trendy clothing items) are sent by truck, due to the pressure to get the product on shelves in a timely manner; “standard” products that are not trendy or seasonal (e.g., paper towels) are shipped by rail. There is also a difference in the mode chosen if the product is perishable or non-perishable. Perishable goods are typically shipped by truck, while products that have a longer lifespan can be consolidated to fill a railcar, which is more cost-effective. R3, a firm that ships perishable items that require temperature-controlled settings, found that the refrigeration on the railcars does not meet com- pany standards. For that reason, R3 tends to use refrigerated trucks. Mention was also made of bulk products, typically sent by rail and barge, if there is access to these modes. S1 stated that for oversized and overweight cargo that is too large for trucks and rail, there may not be any option other than waterways, irrespective of the value of the cargo and the cost of transportation. S5 typically uses rail because the firm mostly deals with volatile and hazardous materials that can be efficiently and safely handled by rail in large volumes. S2 uses rail for bulk liquid cargo of unit size (85 to 110 car shipments) as it is cost-effective, requiring less labor and handling. Seasonal Changes Seasonal changes, due to weather and sales cycles, were identified by five of the eleven inter- viewees as a key influencer of freight mode choice. Four of the five interviewees specified that this is a major factor during the winter months. Inland waterways, in certain locations, are closed during the winter, due to freezing, and winter weather may also result in delays for rail. This

Factors That Influence Freight Mode Choice 49 results in decreased use of rail, intermodal, and barge, and increased use of trucks. C1 experi- ences a decrease in the shipment of agricultural products during the winter, to almost half the typical volume shipped during late spring and summer. R2 noted that there are changes in mode due to changes in sale cycles. The pressure to deliver supplies before “Black Friday”—when retailers offer significant discounts—results in a faster stock turnover that leads to an increase in the use of trucks for faster deliveries. Drayage and Shipment Distance Four of the eleven interviewees indicated that drayage distance influences mode choice. The interviewees indicated that drayage that exceeds the general range of 100 to 300 miles would preclude the use of rail or barges. Three interviewees consider shipment distance to play an important role in mode choice. Various interviewees indicated that there is a break-even point past which another mode should be used. The receivers mentioned that when comparing truck and rail, the break-even point is a length of haul in the range of 600 to 700 miles; if the shipping distance is longer than that, the receivers will try to ship by rail. Rail also becomes an option for long-haul cargo if water transport is not possible. Moreover, other companies aim to use intermodal as much as possible, especially for trips that are between 700 and 900 miles long. S5 mentioned that the trade-off between distance and shipment size decides the mode; truck is preferred for shorter distances and small shipments. Shipment Size The IDIs revealed shipment size as a key factor in mode choice. Four of the interviewees mentioned that shipment size plays an important role in mode decisions. R3 stated that the rail 3PL provider requires a minimum shipment size to book an entire railcar, which is only possible for R3 during its peak season; otherwise, other modes are used to transport the goods. S1 chooses rail primarily because of the size of the materials. Finally, shipment size is also very important for the carriers. When they are transporting only FTL, the firm has to consolidate goods at the terminals before loading the cargo onto the trains. In the opinion of C2, it is not practical to use rail to handle LTL shipments because of the consolidating effort required. One of the major challenges the companies face is filling a railcar at the right time. Other Factors The factors included in this section are those that were mentioned by two or fewer inter- viewees. One such factor was the infrastructure for inland waterways. S3 stated that access to consistent drafts and the condition of locks and dams influence using inland waterways. If the infrastructure is not appropriate, cargo is diverted to either rail or truck. C1 mentioned that the availability of suitable river terminals and loading equipment is crucial for competitive water transportation. Transit time, as an influencer of mode choice, was deemed important as it deter- mines the cost of inventory in transit and when the receiver would get the supplies. This leads to the use of faster modes (such as trucks) rather than intermodal or rail if the company needs the product in stores more quickly. When the companies have flexibility on time, they can utilize the more cost-efficient modes. There is also a concern about damage inflicted on cargo during transportation. R3 told the research team that the firm is not willing to ship delicate products, such as certain fruits, by rail because there is a larger risk of damage. Another factor mentioned by a couple of shippers (S4 and S5) is the influence of receivers/ customers in mode choice. S5 indicated that the client is the one that decides on the mode: “Based on what our customer wants, jointly decided by our commercial, transportation groups and the facility work together on what mode to use.” C2 mentioned that the customer’s ability to track the shipment throughout its journey is a plus for a mode. Another factor was identified

50 Impacts of Policy-Induced Freight Modal Shifts by R2, who said that the need to perform backhauls from DCs to vendors has led to the use of trucks despite his firm’s interest in using intermodal. One of the most interesting comments received was made by a mid-size shipper (S4) that indicated that the firm takes into account the size of the potential transport provider when making mode choice decisions. In the firm’s view, large transport providers (in all modes) may not care much about providing a high quality of service to a much smaller client. In the firm’s experience, transport providers of comparable size to their company provided much better service because the company’s business was more important to these carriers. The firm would not consider dealing directly with the railroads or large trucking companies, which it found insensitive to the needs of mid-size shippers. This perception presents a major hurdle to increas- ing rail market share. It also presents an opportunity to rail-based 3PL providers. These smaller and more agile companies purchase rail capacity in priority lanes, which they proceed to sell to their customers. The terms of their contracts with the railroads ensure that the 3PLs get prior- ity service and relatively short transit times. Moreover, by arranging the drayage to/from actual origins and destinations, these 3PLs provide a quality of service that may compete favorably with that of long-distance trucking. In the research team’s opinion, operations like these provide the best chance of increasing rail usage. Suggestions to Increase Rail Market Share The interviewees were asked to suggest what the public sector could do to increase rail use. Most of the firms indicated a willingness to use more sustainable modes provided certain condi- tions were met. S4 mentioned that some customers ask for the firm’s involvement in sustain- able programs such as SmartWay, before ordering goods. Another large receiver (R2) said that shifting to sustainable modes may reduce the firm’s transportation costs. The answers to this question provide insight into potential policies that may be able to induce mode shift and the probability of acceptance of policies of this nature. Table 11 presents the summary of major suggestions/improvements for each mode and respective quotes from the interviewees. Rail Most shippers and receivers are willing to use rail if the quality of service is up to their stan- dards. R3 indicated that there is a need for better temperature control, reduced damage to sen- sitive goods, and consistent on-time deliveries. S2 expressed willingness to use rail if it offers competitive pricing, consistent service, and availability of labor and equipment. (This firm often experiences delays due to what it perceives as insufficient labor and equipment at railroads.) The shipper also mentioned that an early arrival of goods can lead to storage problems and extra costs. S3 pointed out that the operations of manifest trains—a mix of various railcars and cargo types—are not as well run as those for unit trains. S4 mentioned that the firm does not get proper customer service and cannot track shipments once they are in rail. Despite adopting a rigorous routing plan, C2 experiences delays due to unpredictability in rail schedules, especially when switching between different railroad providers. It is not clear to the firm when arrivals and departures of trains are going to occur. S5 is keen on choosing a better railroad provider as the closest one is not always reliable; this firm also believes that the reciprocal switching regulation proposed by the Surface Transportation Board, in which two or more railroad providers allow each other to operate on lines owned by them (Booth et al. 2016), could influence the future railroad choices. Intermodal S2 expressed a willingness to use intermodal if the savings realized from the rail portion compensates for the costs associated with the extra handling. S2 also said that the firm uses

Factors That Influence Freight Mode Choice 51 Improvement Quotes from the Interviewees Ensure on-time deliveries “We get low 90s on on-time percentages and derailments and what’s happening lately in the industry. How is fracking fuel affecting us? Because rail is making intentional choices to put more of that fuel and oil on trains than there are intermodal stack cars...” Ensure consistent deliveries “Normally it takes 20 days or 15 days transit from my place to my customer. Yesterday or last month you were doing it in 28 [days] instead of 15, and then this month you are doing it in 11. So, how do I plan my inventory? How does my customer plan their inventory?” Improve manifest rail deliveries “They [railroads] move a block of cars from point A to point B very well, 50–75 cars. They also do merchandise manifest business well but not nearly as well as those of block of trains… We do support a competitive rail industry.” Improve customer service “It is a not a very customer-service driven type operation… so that is frustrating to us… But to call them and to get service from them, most times you don’t get any” Efficient switching between railroads “Say we need some cars switched or we need to marry-up some cars, there is a work event on the railiroad sometimes that's quick, sometimes it’s slow, but it is not very transparent exactly when the departure or arrival is going to happen.” Availability of backhaul “We might send 130 railcars from West Coast to East Coast; and then going back right now, it is anywhere between 1 to 7 or 8 [railcars].” More geographic spread “There are areas we would like to see rail service but [it’s] not a good option, by the time we link the dray[age], it’s not always [cost] competitive.” Consistency in transit time “They need to provide the level of service we are looking for.... So, if you tell me its going to be 4 days, hit the 4 days everytime.” Reduce transit time “We are always looking for opportunities to increase intermodal …… The challenge is it has to be fast enough that we want it” Decrease congestion “You know about Chicago. Roughly, I think 25% of domestic rail shipments interchange at some point in and around Chicago. That congestion causes a great deal of issues especially in the winter months.” Locks/dams in good condition “If we want to shift… more tons to water then we got to effectively and consistently maintain our locks and our dams and our waterways.” Dredging of canals “(If) I get an extra two foot; that is 500 tons.....I can increase my tonnage, we can start moving salt…rock... Right now it is probably cheaper to truck them…We could not afford it [dredging] That is just way too beyond.” Preserve land along rivers “We need terminal space... with the ... over-the-road, 500 metric ton crane, all we need is a piece of land... All we need is access... waterfront property has become very expensive, and... gobbled up for housing ... this working kind of waterfront still needs to exist.” Improve terminal operations “Many of them [terminals] have been turned into marinas... They already exist...They just have to satisfy and prioritize some for cargo. You wouldn’t need that many because we also can unload in a dock. We have many many docks, but what they could do is provide some engineering information…what the allowable loading is for the dock.” Mitigate driver shortage “It used to be, you could call a carrier and have anything you need at anytime.... Now, it's just not the same.” Increase the weight limits “Heavy truck legislation, at some point we should get that right…. If there is a safe way to move more than 80,000 pounds across this country we should be looking at that.” Table 11. Suggested improvements.

52 Impacts of Policy-Induced Freight Modal Shifts intermodal for distances longer than 250 miles and trucks for all other shipments. R2 stated that it cannot use intermodal despite its willingness to do so because of rail congestion, especially in the peak season. In R3’s opinion, the intermodal market share could be increased by supporting businesses that provide truck-rail 3PL service, as these firms take care of storage and drayage. The unavailability of intermodal at some geographic locations is forcing R4 to use trucks, which is expensive. This firm also faces problems with inconsistent transit times from intermodal ser- vices at other locations. Inland Waterways S1 mentioned that waterways can be a good choice, especially for larger shipments. S3 is willing to switch from truck to waterways if the maintenance of locks and dams improves. R3 experiences long waiting times, while S4 is not satisfied with the level of service offered by barges. C1 is convinced that the waterway mode could increase its market share, if a few improvements were made. The first is the dredging of local canals, which has not been done for years. The second improvement is to preserve the land along rivers that is being converted into either residential or recreational centers at a fast rate. The last suggestion is to improve terminal operations. The interviewee mentioned that many river terminals are abandoned; it is important to make them operational so that the firm could start using them. Trucking A couple of shippers (S2 and S4) were affected by the driver shortage produced by the changes in hours-of service safety regulations that went into effect on July 1, 2013 (FMCSA 2013). Accord- ing to them, numerous truck drivers are exiting the market because of the working conditions and, particularly, the long stays away from their families. A shipper (S3) and a couple of large receivers (R1 and R4) said that they would encourage an increase in the weight limit for trucks, as doing so reduces the transportation costs and decreases the number of truck trips. Discussion and Conclusion These results are from the IDIs conducted with 11 firms comprised of four receivers, five shippers, and two carriers to gain insights into their mode choice decisions. The research found that most of the firms went through changes in their logistics patterns over time. A shipper mentioned that the firm is doing more on-time deliveries to customers. A carrier has started to provide warehouse and repackaging services for customers. Major factors influenc- ing their mode choices are cost, reliability, consistent transit times, better tracking, product type, value, inventory, and drayage. Most of the shippers and receivers use truck, a mode they find more reliable and faster than rail or intermodal transport that also requires less handling. The major problems for the interviewees with rail or intermodal are inconsistent transit times, poor tracking, handling, and temperature control. Congestion in the intermodal lanes and delays in switching between railroads are also major concerns mentioned by both a shipper and an intermodal carrier. A shipper mentioned facing problems with congestion in inland waterways. When the interviewees were asked about required improvements, the majority of the ship- pers or receivers stated a willingness to switch to rail or intermodal if they could get consistent, reliable deliveries with better customer service. A shipper mentioned wanting more geographic spread in railroads. With regard to waterways, a shipper and a carrier stressed the importance of dredging, as even a few feet of increased depth makes waterways more cost-effective. The importance of preserving land along rivers and regular maintenance of waterway terminals were also mentioned. For trucks, a couple of receivers mentioned that an increase in the weight limit

Factors That Influence Freight Mode Choice 53 would reduce costs by decreasing the number of trips. Overall, participants are willing to switch to more sustainable modes if quality of service requirements are met. From the policy perspec- tive, many improvements would be required to achieve a better modal split. Review of Efforts to Induce Changes in Freight Mode Shares This section provides an overview of policies aimed at inducing mode/vehicle shifts for freight transport at the national and urban levels. National-level policies focus on a range of modal shifts, whereas urban-level policies aim at various truck sizes, as explained below. National-Level Policies Freight mode/vehicle policies are typically designed to combat negative externalities such as congestion, deterioration of infrastructure, air and noise pollution, and accidents. The imple- mentation of freight mode/vehicle policies may also result in externalities and unintended con- sequences. This section discusses relevant policies and their impacts. A few studies concentrate on the factors that influence shippers or carriers to shift from road to rail mode. Policies that may not be intended to cause mode shift may inadvertently cause a shift from a more sustainable mode to one that causes more emissions and environmental impacts. For example, the British government set the goal of increasing rail mode share by 80 percent by the year 2010, as a means of reducing the negative impacts caused by road trans- port of freight. The need for this intervention arose due to policies that had the opposite effect, such as improvements in roads, lower fuel costs, and fewer investments in rail and other sustain- able modes. The reasons for the preference for road in freight as identified by Woodburn (2003) are pressure from customers for JIT delivery (which causes low-volume flow), road accessibility to all locations, poor level of service offered by rail to customers, and lack of funding for new and improved rail infrastructure. It was also found that rail could serve as a potential mode for 37 percent of freight in 5 years, provided supply issues were addressed (Woodburn 2003). Dewey et al. (2002) studied the potential of subsidy and tax policies to correct the market’s misallocation of freight shipments between the trucking and railroad industries, the existing equilibrium in the surface freight transportation market, and the optimal subsidy. With a sensitivity analysis, Dewey et al. concluded that subsidizing the shipment of freight by rail- road reduced truck shipments and improved economic efficiency by more than enough to offset the loss in efficiency caused by collecting the revenues to pay the subsidy. Stewart et al. (2008) studied the impact of rail-to-truck mode shifts on pavement structures in three major corridors in the United States. The mode shift to truck is often due to the closure of short-line railways and growth in economic activities (new industries) in an area. Asset management techniques are used to estimate the cost of damage to pavement due to increased truck traffic. Stewart et al. estimated that 12 out of 32 pavement sections in the Wisconsin and South Railway line and 18 out of 48 pavement sections in Michigan face reductions in life- span. The viability of modes is also addressed in the study. In the case of the Minnesota Prairie Line closure, the new ethanol plants increased truck volume by 400 percent, making 11 out of 14 pavement sections inadequate to accommodate such traffic. In addition, rail cannot be replaced by trucks due to high volumes and the volatility of the ethanol. McAuley (2010) evaluated the cost of externalities due to rail and road freight on four major corridors in Australia, showing that the rail mode causes fewer externalities than road. The externalities evaluated are shown in Table 12. The table gives the costs of externalities in cents

54 Impacts of Policy-Induced Freight Modal Shifts per ton-kilometer (ton-km) for rail and road freight movement. The total externalities per ton-km are greater for the road mode than they are for the rail mode. For both road and rail freight, the major externalities are accidents and GHG emissions. However, the cost of accidents and GHG emissions for rail freight is much lower than it is for road freight. This study concludes that road freight is underpriced relative to rail due to externalities and suggests a mode shift to rail as one way to mitigate negative externalities. Gleason et al. (2005) evaluate the net benefits associated with increasing the utilization of rail mode from 25 percent to 100 percent of its capacity. The benefits include a reduction in the costs of such externalities as congestion, accidents, pollution, noise, and infrastructure damage, and the costs associated with fuel savings. Utilization of rail at 100 percent of capacity increases the freight rail mode from 14 percent to 56 percent. Gleason et al. (2005) finds that the major reasons for underutilization of rail capacity are (1) the weight ceiling is less than other lines, and (2) there are limited transloading facilities. The major benefit in terms of reduction in CO2 emis- sions would be $388,000 every year. The total net benefits are evaluated as $13.5 million every year. This study also evaluates the boost to the economy in terms of employment; every million dollars of revenue from rail annually creates 11.4 new jobs. The transportation of hazardous materials by rail is more efficient, which is not considered in Gleason et al. (2005). Beuthe et al. (2002) studied the benefits associated with the new modal split occurring after internalizing the external costs of road users by changing pricing policy. This work evaluated the externalities of congestion, pollution, accidents, noise, and damages to infrastructure and made the users responsible for these externalities. With the new pricing, the modal split was simulated using NODUS virtual network methodology (a systematic GIS designed to analyze freight trans- portation over long distances), optimizing generalized costs over links on a virtual network. The results showed a decrease in congestion costs of 44 percent, pollution costs of 14 percent, accident costs of 24 percent, noise pollution costs of 20 percent, and damages in the 1995 traffic scenario of 27 percent. Modal shift is highly effective in reducing external costs to society. Wiederkehr et al. (2004) determined the definition, criteria, potential policies, and their effects for an Environmentally Sustainable Transport (EST) project initiated by Organization for Economic Co-operation and Development (OECD) countries. Based on case studies from nine OECD countries, Wiederkehr et al. (2004) concluded that 40 percent of the EST project can be achieved through technological improvements, while the remaining 60 percent should be achieved from demand-side changes and modal shifts to sustainable modes. The criteria for EST are based on the emissions of CO2, NOx, VOC, and particulate matter, as well as noise and land use. The emissions of CO2 from transportation should be 20 percent to 50 percent of such emissions in 1990, and emissions of NOx and VOC should be less than 10 percent of the emissions in 1990. The PM-10 should be Road Rail Road Rail Road Rail Road Rail Accidents 0.060 0.080 0.260 0.080 0.210 0.080 0.120 0.080 GHG 0.104 0.054 0.090 0.054 0.100 0.054 0.100 0.054 Noxious emissions 0.018 0.010 0.018 0.005 0.005 0.008 0.003 0.003 Noise 0.001 0.003 0.001 0.004 0.001 0.003 0.002 0.001 Congestion 0.022 0.000 0.014 0.000 0.011 0.000 0.005 0.004 Total 0.205 0.147 0.382 0.143 0.327 0.145 0.230 0.142 Externalities Corridor-1 Corridor-2 Corridor-3 Corridor-4 Source: McAuley (2010) Note: Damage to pavements or rails is not considered in the study. Table 12. Summary of external costs of freight movement 2010 (cents per ton-km).

Factors That Influence Freight Mode Choice 55 reduced to 55 to 99 percent of the 1990 levels. The noise levels should not be more than 55 deci- bels (dB) during the day and 45 dB at night. Land use should meet regional objectives for air, water, and ecosystem protection. The primary solution to meeting the criteria listed in the previous paragraph is switching from less sustainable to more sustainable modes, accompanied by a relative decrease in the unsustainable modes. This goal is called the modal split of EST and is planned for 2030. The goals of the modal split of EST in 2030 include 60 percent of freight ton-km transported by rail or combined modes, and 20 percent of freight ton-km transported by waterways. In addition, rail and waterways should incorporate increases in capacity, speed, and use of more sustainable fuel sources, such as electric or hydrogen-cell-based fuel. Even the production of electricity has to be made efficient, and the electricity has to be made mostly from renewable energy sources. This change requires targeted investments in sustainable modes and energy, combined with policy regulations. The modal split of EST in 2030, compared with projections from the current modal trends and externalities have been evaluated up to 2015. Accidents are the major externality in both cases. In EST, the total cost of externalities except congestion is reduced by 37 percent by 2015. In 2015, for nine countries, the cost of externalities is 5 percent of the GDP in the current scenario and 2 percent of the GDP in EST. Modal shifts alone could cause a dramatic decrease in external costs. The current cost of externalities, which is evaluated as 6 percent of the GDP, is due to past modal shifts that caused current modal patterns. The modal split of EST in 2030 has its own negative impacts on the economy, such as reductions in GDP growth and the employment rate, but compared to the estimated reduction in externalities these negative impacts are minimal. Urban Freight Policies One type of policy that has had unintended consequences is truck size restrictions, or the banning of larger trucks from urban centers. These types of policies are implemented with the aim of decreasing congestion, which in turn results in decreased pollution and increased safety for road users. However, the implementation of this policy results in larger shipments that would typically be delivered in a large truck, being broken down into smaller shipment sizes and delivered in smaller trucks, to comply with the regulation. The policy therefore results in more than one small truck replacing a large truck to compensate for the reduction in capacity. This concept was explored in research conducted by Holguín-Veras et al. (2011), which com- pared the performance of large and small trucks delivering in urban areas, in relation to the social costs produced. The results indicated that when using social cost as the determinant for which truck class is optimal, it is essential to not only analyze the amount of social cost generated, but also to include the cargo productivity and the substitution effect between the various classes. Using this method, the analysis was able to capture the true response to restricting larger trucks. The amount of cargo will not be reduced by limiting the truck traffic, as this will have negative effects on the economy; therefore, what will occur is the replacement of a larger truck with a number of small trucks that will be able to cover the amount of cargo that is being delivered. Even though the aim of restricting larger trucks in urban areas is to reduce congestion and the negative externalities resulting from congestion, the net effect may be the opposite, and actually increase social costs due to an increase in VMT by small trucks. Using the Oakland, California, network as a case study, a combination of traffic micro- simulation, econometric modeling, optimization techniques, and valuation for the externalities was used to determine the optimal mix of traffic and the isocost substitution rate, which was complemented with a sensitivity analysis to evaluate the effects on social cost. Application of

56 Impacts of Policy-Induced Freight Modal Shifts the optimization formulation to the base-case scenario indicated that the use of large trucks only would provide the optimal traffic mix that would minimize social costs. The use of large trucks only when required by the shipment results in social cost savings of 2.76 percent, a reduction in average congestion time by 3.84 percent or $76,151/day, and reductions in various pollutants from 0.02 percent to 8.68 percent (Holguín-Veras et al. 2011). The sensitivity analysis, using various combinations of small and large truck payloads, indi- cated that the most crucial factor contributing to social costs is the substitution rate between small and large trucks, which was expected, based on theory. The isocost substitution rate was found to be 1.6. In other words, if the average payload for a small truck is less than 62.5 percent (1/1.6) of the average payload of a large truck, then the use of large trucks would result in a reduction in social costs. If this is not the case, then small trucks are the better choice. The results of the study are important for policymaking because they combat the assumption that minimiz- ing large truck traffic will result in a decrease in social costs without factoring in the substitution effects. Therefore, the take-away for government and planning agencies is to assess these substitu- tion effects to have a broader, more conclusive view of social costs during the decision-making process (Holguín-Veras et al. 2011).