Computer Assisted Cost Assessment of
Intermodal Transportation Linkages
Phase I

Final Report on MBTC Project #1020

Submitted to

The Mack-Blackwell Transportation Center
University of Arkansas

July 26, 1995

Endorsements:

⁴4a^p_"AOUL Eric M. Malstrom Darlene P. Butler Principal Investigator Co-Principal Investigator Professor and Head Assistant Professor

Raja Kasilingam Co-Principal Investigator Assistant Professor

Table of Contents

Background 1 II Objectives 4 -III Regional Definition 5 IV Accomplishments to Date 7 V Literature Review 9 VI Method of Analysis 47 VII Results of Analysis 84 VIII Development of Transportation Cost Software 89 -XI Summary 106 IX References 107

I. Background The most fundamental characteristic of economic society is trade. Transportation is

^{inseparable from trade}. Transportation, more than anything else, determines the "extent of the

market." Economic progress would be virtually impossible without the ability to carry goods

^{from place to place efficiently and cheaply}. Distance has ceased to be an obstacle to trade or

^travel. Needless to say, the improvement in transportation, which has made possible the modem

system of commerce, has brought about a vast improvement in the standard of living of all people. These improvements have provided a volume and a diversity of goods for individual

consumption which were quite unattainable in former years, when the expense and the difficulty of transportation were such that nearly all trade was purely local in nature.

For the past few decades, very few studies have been done to compare the cost of the various transportation modes . Historically, comparisons have been based on the cost to transport a given unit of weight or volume per unit of distance traveled . Earlier studies have been done primarily to determine rates in the transportation industry . Virtually none of them address the societal costs of providing transportation services. Several government entities contribute to the nation's transportation infrastructure making it difficult to comparatively assess the true cost of different transportation modes.

The rail industry, for example, incurs costs to purchase right of ways, install and maintain track, and to construct bridges, tunnels, crossing signals, etc . . Railroad operators have most of the financial responsibility for maintaining the railroads' facilities, but the rates of rail transportation have long been regulated by the federal government . In contrast, trucking companies transport goods via the nation's highway system . The trucking companies are not

responsible for the construction of their transportation network, but instead they pay an annual tax for each truck that uses the nation's interstate and local highway systems .

Waterway transportation has long been regarded as an economical way to transport bulk goods. It consists of the huge system of locks and dams of this nation's navigable inland waterways. The Army Corps of Engineers, a federal government agency, has financial responsibility for the waterways. A similar situation exists with air freight transportation. Air freight companies do not pay for the construction of the air fields and air ports . These facilities are funded by the federal and state governments.

The expenditures by state and local governments include federal grant-in-aid funds while

the federal expenditures are generally direct outlays . Table 1 presents the ratio of transportation expenditures that are "covered" or paid for directly by users and transportation-related collections.

Table 1: Federal, State and Local Government Transportation Finances (millions of dollars) [611

II. Objectives The investigators are developing a cost model to determine the societal cost associated with different transportation modes. The different transportation modes used in this analysis include rail, truck, barge and air transportation. The cost model will consider the costs incurred by all activities associated with each transportation mode. These costs will be amortized to result in a true (societal) weight/volume per unit distance cost for each transportation mode . These costs are generally not reflected in the transportation rates associated with each mode, even though they are a real and tangible burden to taxpayers . The obtained costs for each transportation mode from the proposed cost model will yield comparatively true values of weight/volume per unit distance costs. These obtained costs are representative of the societal costs of providing transportation services. The developed model is being implemented in a PC/Windows based software . The developed cost model will serve as the basic structure for the software development effort being

completed in the second phase of this project. This software will enable the user to determine the most effective transportation mode between a given origin and destination .

III. Regional Definition

The region selected for this analysis is primarily a rural area. The selected region includes five states, which are Alabama, Arkansas, Mississippi, Louisiana, and Tennessee . Four of these states were selected because they are contiguous with Arkansas. Alabama was added to the analysis region because of its ocean coastline associated with water transportation .

A. Truck Transportation Network

In order to keep the roadway transportation network of manageable size, only interstate highways within the region are being considered. This selection was also motivated by the availability of cost data for interstate highways as opposed to state and secondary roadways . The interstate highways in the selected region include 1-40, 1-20, 1-55, 1-59, 1-10, and 1-65 .

B. Waterway Network

Most of the U.S. waterway networks are concentrated both on the Great Lakes and the Mississippi River Basin. The region selected for analysis has an extensive network of rivers and canals. The river systems included in this analysis are the Lower Mississippi River, the McClellan-Kerr Arkansas River, the Black Warrior-Tombigbee Rivers, the Tennessee River, and the Alabama-Coosa Rivers .

C. Railway Network

This selected analysis region is served by numerous railroad operators . These operators are nationwide providers of railroad freight services . These operators include Burlington Northern Railroad, CSX Rail Transport, Kansas City Southern Line, Norfolk Southern Corp., and Union Pacific Railroad. Only the current ICC Class I railroad companies are included in this analysis.

This is an Interstate Commerce Commission (ICC) classification which includes railroads with average annual operating revenues of $50M or more.

D. Air Transportation Network

In Arkansas alone, there are over 100 airports and airfields of many kinds . To keep the size of the network to a manageable size, only airports that are classified as primary and commercial under the National Plan of Integrated Airport Systems are included . This reduces the number of airports in the entire region of analysis to about 40 .

IV. Accomplishments to Date

This project is being carried out in two phases as it has been proposed to the MBTC . Phase I is being completed currently over an eighteen month period . Phase I consists of two major tasks, development of a cost model, and implementation of the model in PC/Windows based software. We have focused on the development of a cost methodology to assess the various components of the societal costs associated with each transportation mode being analyzed . The

methodology is common to all transportation modes . It permits each transportation mode to be accurately compared without the biases of federal/state contributions or intangible costs. We are also working on developing a software to determine the true weight/volume per unit distance costs associated with different transportation modes .

The second project phase (Phase II) is scheduled over a twelve month period, and it began in August 1994. Phase II will focus on the application and use of the cost methodology and software developed in Phase I. In this phase, the software is being enhanced to provide the user with a method of selecting the least cost transportation mode for shipping from a selected origin to a desired destination within the analysis region. Our accomplishments since beginning of this project are summarized in the following sections .

A. Literature Search

An extensive literature search has been conducted to obtain any data and information that could be beneficial to this research . As of today, there are over sixty viable sources cited in the bibliography. We have noted that very little prior research has been conducted in recent years to address transportation cost allocation .

B. Cost Model Development

To date, a general cost model has been developed to assess the true cost of transportation .

Due to the different cost categories associated with different transportation modes, each model

^{was tailored to fit each transportation mode} . Each individual model was derived from a general

cost model. Data collection represents the most difficult task in this research in terms of the data availability. Data from the federal, state, and ICC regulated trucking companies were collected for the analysis of truck transportation. Data from ICC regulated railroad companies were

collected for the analysis of railroad transportation . Data from the federal and state government,

and data from the airlines compiled by Department of Transportation's Office of Aviation were

collected for analyzing air transportation . Data needed for waterway transportation analysis

were gathered through the Institute of Water Resources. For the purposes of consistency of analysis, cost data for fiscal years 1989, 1990, and 1991 were collected and used in our analysis, with the exception of waterway transportation.

C. Software Development

The software has been written and developed in Windows environment. It also provides a user-friendly interface to run and iteract with the cost model . The software can evaluate single or multi-modal routes as well as evaluate alternate routings between an origin and destination . A

sample session of the software will be provided in the final project report . A detailed description of the software is provided in Section VIII .

V. Literature Review

The goal in conducting the literature search has been to obtain data and information that

could be used in assessing different transportation modes. To date, a bibliography in excess of

^{sixty viable sources has been compiled} . Perhaps the most striking results of this exercise is the almost complete absence of any consistent cost data for the different transportation modes . We are attempting to assess the federal and state contributions to the transportation infrastructure of each mode being analyzed, and the operating costs incurred by each transportation carriers for each mode.

Cost related transportation studies have been devoted to rate setting by the transportation service providers, and determination of tax structures by state and federal government . Purchasers of transportation services usually consider the cost incurred to them rather than the cost incurred by the society in general .

A. Highway Transportation Literature on this transportation mode is summarized in the subsections that follow .

1. Truck Usage and TruckMiles of Travel

Approximately 44.6M private and commercial trucks were registered in the United States in 1987. Two-thirds of all trucks were used mainly for personal transportation . Alabama had about 894,000 trucks in use. The average miles per truck was 12,400, and the truck miles

traveled was 11B in Alabama in 1987. There were about 521,000 trucks in Arkansas in the same

year. The average miles per truck was 11,200, and the truck miles traveled totaled 5 _.8B. Louisiana had more than 926,000 trucks in service, and the average miles per truck traveled was 12,200. The truck miles driven in Louisiana was about 11 .4B. Mississippi had the lowest number of trucks among the five states with slightly over 500,000 . The average miles per truck is 11,900 and the truck miles driven totaled almost 6B . Tennessee had the largest number of trucks in these five states, 1 .02M. It also had the highest number of average miles per truck,

12,500. The truck miles driven in Tennessee is 12.7B [51] .

The average vehicle-miles of travel by all trucks on interstate rural highways was 69 .4B, and 75.2B vehicle-miles on interstate urban in 1992 [51] . The total number of vehicle-miles by all trucks on all roads was 629B . Vehicle-miles is defined by Federal Highway Administration

(FHWA) as the product of total mileage traveled by all trucks .

2. Highway Cost Allocation Studies Highway cost allocation requires addressing the problem of how to raise revenues from

selected groups of taxpayers so as to meet a given budget in a fair and equitable manner [33] . Efficiency and equity are the two fundamental standards against which to evaluate user-charge instruments and rates [31]. Lee [28] noted that efficiency is not mentioned in the congressional mandate for the federal highway cost-allocation study now under-way . The notion that vehicles

should pay for the costs they generate is described as equitable, but it is also essential that they do so because it encourages them to reduce these costs and make sure that the benefits they derive are greater than the costs created. Efficiency can be divided into short-term efficiency and long-term efficiency. Short-term efficiency deals with fording ways to maximize the utilization of the transportation facilities, while long-term efficiency deals with finding the best program of investment in fixed facilities [28] _{. The concern addressed by equity is the}

distribution of costs and benefits among groups within society . Horizontal equity is not directly

related to popular ideas of fairness ; it urges that equals be treated equally . Vehicles in equal circumstances, from the standpoint of the highway provider, should therefore be charged equally.

The allocation of highway costs among the various classes of highway users is an area that has received an increasing amount of attention with the completion of the new Federal Highway Cost Allocation Study and several state studies [51] . Cost allocation distribution is a fair way to

allocate costs among users. The amount of cost generated by a vehicle class can be determined by desegregating items of expenditure and assigning them to vehicle classes [28] . An algorithm known as the incremental cost method has been popular in recent studies, but there is very little professional consensus on which are the best methods for the practical determination of highway-user cost responsibilities. The incremental cost method takes as its starting point a basic highway, usually one designed for automobiles alone. Additional classes of vehicles cause additional increments of cost, and these increments are apportioned among the members of each class.

In addition to capital and maintenance expenditures, the following costs should be included

[28] .

Hidden costs: Some costs appear in public budgets but not in the budget of the agency responsible for highway expenditures. Vehicle code enforcement and traffic control may be hidden in police budgets, and payroll administration may be centralized rather than included in the transportation agency budget.

Negative: Negative externalities in the form of air pollution, noise, water pollution, and other unpriced effects on the physical, natural, and human environment constitute real costs to society. Even though we may never be able to place accurate dollar values on these costs, present policies can be improved on by a recognition of such costs .

Interference: Private costs in the form of delay time, vehicle wear, fuel, and accident costs are relevant to the correct pricing of highway services .

Tax expen-: Exemption of fuel from general sales taxes and exemption of highway

ditures property from local property taxes result in a favorable treatment of highways in comparison with other activities that are not exempt . To the extent that these taxes pay for general government services, highway users are being subsidized by those engaged in other activities .

Interest: The pay-as-you-go philosophy, in which each year's expenditures are foregone matched with the same year's revenues, hides the fact that invested capital has an opportunity cost represented by the rate of return that the money would earn in another activity .

Theory tells us that the price charged for use of the highways should be equal to the marginal use cost. If less is charged, the user may not value the use as much as society values the resources consumed. If more is charged, some potential users are deterred, even though they would gain more from the travel than it costs society . However, this principle only applies to charges, such as a fuel tax or a weight-distance tax, that vary directly with usage . The design of practical mechanisms for imposing correct prices on highway users presents a major challenge .

Selection of a single budget for cost recovery from the many government agencies that participate in financing the highway system is highly arbitrary . The highway infrastructure lacks a single authority that can establish a complete user-charge structure . Without this institution, the only possible surrogate is a federal government initiative [3].

In allocating highway costs among vehicles, the two primary considerations are vehicle mileage and vehicle weight . Costs that are considered to be common to all vehicles are allocated to vehicles according to mileage [3] . Vehicle weight must also be taken into account when allocating cost responsibilities because the highway damage occasioned by a vehicle increases

geometrically with the weight it carries on each axle . This means that highway costs are

affected by the number of axles on a vehicle, the configuration of the vehicle's axle, and the

distribution of its weight on the axle [3]. Data from the Federal Highway Cost Allocation Study

[51] shows that the average passenger vehicle had a cost responsibility of $3 .27 per thousand miles traveled, while a combination truck weighing over 75,000 pounds has a cost responsibility of $47.14 per thousand miles traveled [3]. A Highway Cost Allocation Study was conducted in 1981 by the Colorado State Highway

Commission [51]. Two vehicle classes were utilized in the study, the two classes were defined as "basic vehicles" (weighing 10,000 lb. or less) and "heavy vehicles" (weighing over 10,000 lb.). In this analysis, cost responsibility is considered to represent the relative share of the Highway Users Tax Fund (HUTF) each of the vehicle classes should pay . The overall results show that

heavy vehicles contribute 23% of the HUTF while being responsible for 37% of the HUTF expenditures.

In 1980, the Oregon Department of Transportation conducted a cost responsibility study on two groups of highway users, the basic vehicles and the heavy vehicles . The results indicated that basic vehicles (weighing 8,000 lb . or less) should contribute approximately 55% of the road user revenue, and heavy vehicles (weighing over 8,000 lb .) should contribute 45% [51].

Sinha and Fwa [43] indicated that in a cost allocation study of highway expenditures, allocation of highway pavement costs is the area that is most amenable to technical analysis . Consequently, a certain degree of subjective judgment is usually involved in allocating pavement costs to vehicle class. Over the years, several cost allocation studies have been performed for pavement cost responsibilities, and these studies can be categorized into three different approaches. These include the traditional incremental approach, the direct allocation approach,

and the damage function approach . A unified approach for allocating pavement costs by Sinha was used in the 1983-1984 Indiana Cost Allocation study [43]. The latest Federal Cost Allocation Study [61 ] used a uniform removal technique in its analysis.

3. Cost Functions

A translog cost function presented by Harmatuck [18] was used to approximate the actual cost function of motor carrier related to service quality and marketing improvements . The basic cost model relates carrier operating cost less depreciation (C) to output levels and quality characteristics (Y), and input prices (P) . This translog cost function has the following specification:

i _-i Pi Piit + (1/2) Ei Ek aik Y;it Yikt + (1/2) Ei Ek Rik Put Pild + Ei Ek Sik

Cit + ao+1iaYiit+7

YiitPikt+EiµiDiit+ Eit ⁽¹⁾

where, _Cit = natural log of operating expenses less depreciation of carrier I in period t _Yiit-natural log ofTLshipments of carrier i in period t

Yizt-natural log ofLTLshipments of carrier i in period t Yi3t-natural log ofTL average shipment size (ton per shipment) of carrier i in period t Yi4t= natural log ofLTL average shipment size (ton per shipment) of carrier i in

period t
_Yist - natural log of average length of haul carrier i in period t
Pilt -natural log of line haul activity price of carrier i in period t

Pi2t = natural log of pickup and delivery activity price of carrier i in period t

Pi3t = natural log of other prices of carrier i in period t and a1, /3j,(3jk,6jk,andµjare parameters. The cost function is specified in terms of activity prices rather than natural factor prices . This model was used to compare the service quality and price difference among motor carriers .

4. Current User Charge Structure Annual user charge payments for automobiles have been decreasing . In contrast, these

payments for trucks are increasing. This is primarily because the user charge structure relies heavily on fuel taxes, and auto fuel efficiencies are improving at a faster rate than truck fuel efficiencies. In addition, the truck sales taxes are ad-valorem taxes, and the receipts rise with inflation [61] .

While the tax payments of all trucks as a class were found to be commensurate with the

costs occasioned by trucks, lighter trucks were found to be overpaying by anywhere from 25 to 200% [3]. It is found that the heavier single unit trucks are found to significantly overpay relative to other vehicles . Combination trucks are found to underpay relative to other vehicles . Heavier combination trucks significantly underpay while the lighter combination trucks overpay .

The excise tax on new trucks is a flat 10 percent of the rate on all trucks greater than 10,000 pounds gross vehicle weight (GVW) . The truck parts and accessories tax on all trucks is a flat 8 percent of the price. The highway use tax is a flat $3 per thousand pounds of GVW for all trucks over 26,000 pounds . The truck user charges are not graduated by weight, whereas cost assignments increase with weight due to the nature of the costs occasioned by increasing the size and weight of vehicles.

The federal government and the state governments collected almost $50B in a variety of transportation user taxes and fees in 1991 [47]. A significant portion of the user taxes are used for federal budget deficit reduction . In 1991, federal tax on the large truck use tax was $575M

[47]

. Taxes collected on trucks/buses/trailers amounted to $1 .05 B[47].

B.

Waterway Transportation Literature on this transportation mode is summarized in the subsections that follow .

1. Inland Waterway System

Today some 170 major commercial ports on the nation's coastlines, river, lakes and canals serve as centers of commerce and growth . Every major metropolitan region either is a port or is closely linked by rail or highway with a port . Arkansas' major export is agricultural products . These products are transported by trucks or rail to the Mississippi , Arkansas, or White Rivers and are then barged to major export markets, such as the Port of New Orleans, for shipment overseas. The U.S. Inland Waterways has a total length of 25,543 miles [4] .

The Mississippi River System alone has over 8,954 miles of navigable waterway . This mileage includes waterways improved by the Federal government, other agencies, and those that have not been improved but are usable for commercial navigation [4]. The Mississippi River System represents nearly 40% of the inland waterborne commerce in the United States [4]. The Mississippi River originates in northern Minnesota and flows in a southerly direction to the Gulf of Mexico. Total mileage on the river is 2,360 miles, and only 1,837 miles of it is navigable

[66] . The annual cost of operation and maintenance of this system is $22M. The navigation season from Minneapolis to the mouth of the Missouri River is from the end of March to the first

week of December. The navigation season is year round from the mouth of Missouri River to the head ofPasses in Louisiana.

The McClellan-Kerr Arkansas River navigation system has a total navigation mileage of 448 miles [66] . It runs from the White River and the Arkansas River to the Verdigris River and to Catoosa, Oklahoma. The annual operation and maintenance cost is $5.3M. The navigation season is twelve months .

The Black Warrior-Tombigbee waterway lies wholly within the state of Alabama and is made up of the Black Warrior, Warrior, and Tombigbee Rivers. The total mileage on the waterway is 466 miles, with annual operation and maintenance costs of $2.5M [66]. The navigation season is 12 months per year.

The Tennessee River is formed at Knoxville by the junction of the Holston and French Rivers. It flows south and southwest into the Ohio river. It has a navigational mileage of 652 miles, and the annual cost for operation and maintenance cost is $1 .8M [66] . The navigation season is 12 months.

The main streams of the Alabama-Coosa Rivers are Coosa, Tallapoosa, and the Alabama Rivers. The total mileage on the waterway is 305 miles [66] . The yearly operation and maintenance cost is $720,000, with a navigation season of 12 months.

2. Expenditures for Commercial Navigation

The total annual construction cost of waterways on the Mississippi River System is $223M . The Mississippi River and its tributaries are allocated $21 .6M, advanced engineering and design are allocated $2.7M, channel improvements are allocated $ 27.5M, locks and dams are allocated $129.6M, and $41.5M is a multipurpose allocation. The total annual operation and maintenance ^{cost on the system is $155}.9M. Cost of operation and maintenance on the Mississippi River and

tributaries is $1 OM, $45.2M goes to channel improvements, $87.7M to locks and dams, and

$13M for multipurpose applications . This makes the total annual construction and operation /

maintenance cost to $378.8M [68] .

Annual operation and maintenance cost per ton-mile on Arkansas-Verdigris Waterway is $24.3M. The operation and maintenance cost is $1.04M on the Cumberland River, $0.27M on Gulf Intracoastal Waterway, $0.13M on Lower Mississippi River, $0.55M on Mobile-

Tombigbee-Warrior Waterway, $16.2M on Ouachita-Black Rivers, and $0.58M on Tennessee River [66] . The levy per ton to recover 100% operation and maintenance cost on the Mississippi River is $0.1212 [34] .

3. US. Ports

Traditionally, the U.S. port system has been developed on the basis of a partnership between local and state port authorities and the federal government . Landside cargo handling facilities and associated infrastructure including prescribed components of federal navigation projects, are provided by local, state and private interests . Navigational capabilities have been the responsibility of the federal government.

For their part, the local, state and private interests have responded to the dynamics of change. In its 1984 report on the status of U .S. public ports, the U.S. Maritime Administration states that local, state and private entities invested some $5B in terminal facilities in the period 1946-1980 and are likely to invest $5B more in this decade . All of this money has been and is being invested in a good-faith reliance upon the federal government to fulfill its traditional responsibility in providing appropriate development and adequate maintenance of the nation's

deepcraft navigation system [50].

Federal expenditures for harbors and port development have been historically financed from general revenues. No user taxes or fees have been imposed for these specific expenditures. In June 1985, Reagan Administration announced an agreement with the Senate Republican

leadership of a proposal for a 0 .04 percent (4 mils, or 4 cents per $100) ad valorem excise tax on cargo loaded and unloaded at U.S. harbors to recover up to 40 percent of Corps of Engineers harbor operations and maintenance expenditures . Monies raised by this new tax would be deposited in a newly established trust fund for such expenditures . This tax would be in addition to certain cost-sharing requirements for non-federal contributions to project costs [50] .

In general, federal expenditures for construction, operation, and maintenance costs of U .S . waterways have been financed from general revenues, rather than from fees or taxes imposed on navigation users. In the Inland Waterways Revenue Act of 1978, however, Congress imposed an inland waterways fuel excise tax, and provided for transfer of these tax revenues to an Inland Waterways Trust Fund [50]. Amounts in the Trust Fund are available, as provided by Authorization and Appropriation Acts, for making construction and rehabilitation expenditures for navigation on the specified waterways the commercial use of which is subject to the fuel excise tax [50] .

4. Costs Incurred by Travel on Waterways

Four types of costs are incurred during or prior to actual travel on the waterways : (1) public capital and operating-and-maintenance (O&M) costs of constructing and operating the waterway; (2) opportunity costs in terms of foregone alternative outputs of the multiple purpose

waterway system (e.g. use of the water for power or irrigation) ; (3) private "straight-through"

operating costs, which would be incurred by tows if no congestion were present (if barges could proceed without delay in carrying out their itineraries, that is all channels and locks were made ready to receive a tow upon arrival); (4) congestion costs [22] .

5. Freight Traffic in Waterways The following data was extracted according to each district from material published by the

U.S. Army Corps. of Engineers Districts [52] . The data is summarized in the subsections that follow.

a. Vickburg, MS, District

The district includes the mouth of the Black River to Camden, Ark . (336 miles), from Mississippi River to the junction of Old and Yazoo Rivers (9.3 miles), from the Old River to the mouth of Yalobusha River (161 miles), and the mouth at the junction of Old and Atchafalaya Rivers, LA., to Fulton, Ark. (455.6miles). Tons of freight transported by calendar year are listed below.

1982 -- 3,619,238 tons 1983 --3,572,416 tons 1984 -- 4,760,933 tons 1985 -- 5,518,947 tons 1986 -- 5,553,568 tons

b. Memphis, TN, District

The district includes the junction with the Arkansas Post Canal to Batesville, Ark. (285.5 miles), and mouth of the Wolf River to Mile 3 of the Wolf River . Freight volume, by calendar year is summarized below.

1982 -- 1,667,120 tons 1983 -- 1,396,566 tons 1984 -- 1,939,063 tons 1985 -- 2,117,009 tons 1986 -- 2,014,757 tons

c. Little Rock, AR, District

This district includes the lower 10 miles of the White River, the Arkansas Post Canal, the Arkansas River between Arkansas Post Canal and Muskogee, Okla., and the Verdigris River between Muskogee and Catoosa, Okla.(445 miles), and the lower 5 miles of Lake Langhofer, and the lower 10 miles of San Bois Creek (17 miles). Freight volumes are again listed below .

1982 -- 7,823,228 tons 1983 --7,567,986 tons 1984 -- 8,521,310 tons 1985 -- 7,725,468 tons 1986 --8,395,856 tons

ci Nashville, TN, District

This district includes the mouth of Cumberland River to Mile 552, the mouth of Tennessee

River to Knoxville, Tenn. (652.2 miles), the Barkley Canal connecting the Barkley Reservoir and the Kentucky Reservoir, (1 .75 miles), and the Clinch River (63 miles) . Freight volumes for this district are listed below.

1982 -- 39,698,464 tons 1983 -- 42,983,396 tons 1984 -- 52,089,222 tons 1985 -- 57,502,620 tons 1986 -- 75,544,351 tons

e. Mobile, AL, District

This district includes the mouth of the Mobile River, bay, and river channels into Chickasaw and Three Mile Creeks, the mouth of the Alabama-Coosa River to Wetumpka, Ala ., the mouth of the Black Warrior and Tombigbee Rivers on Chickasaw Creek to Mile 429 .6 of the Chickasaw Creek, the Tennessee-Tombigbee Waterway(233 .7 miles), the lower 4 miles of the Dog River and the lower 6.8 miles of the Pascagoula River. Freight volumes are listed below.

1982 -- 69,146,334 tons 1983 --67,300,393 tons 1984 --83,425,833 tons 1985 -- 81,800,546 tons 1986 -- 86,879,795 tons

f New Orleans, LA, District

This district includes the river from New Orleans to the Gulf of Mexico (75 .4 miles), the Gulf Intracoastal Waterway to Gulf of Mexico (40.2 miles), the Intersection of Gulf Intracoastal Waterway at Houma to Gulf of Mexico (36 .6 miles), the Atchafalaya River from Morgan City, LA. to Gulf of Mexico, the Old River lock at Mississippi River through Old River to Morgan City, LA. (121 miles), the Gulf Intercoastal Waterway from Morgan City to Port Allen (64 .1 miles), the Calcasieu River to the Sabine River (24 .9 miles), and Port of New Orleans, LA . from Mile 127 to the mouth of the Passes . Freight volumes, by year, are listed below.

1982 -- 205,081,030 tons 1983 -- 162,699,336 tons 1984 -- 166,261,389 tons 1985 --168,920,588 tons 1986 -- 177,233,374 tons

6. Waterway User Charges

For many years, developments in government policy have focused attention on the "free use" of U. S. inland waterways. This has culminated in various proposals for imposing a user charge on the nation's towing industry . The four likely forms of user charges are (a) fuel taxes,

(b) lockage fees, (c) segment tolls, and (d) licensing of floating equipment [34] .

The fuel tax is imposed on diesel and other liquid fuels used by commercial cargo vessels on twenty-six designated inland or intracoastal waterways of the United States (Code sec .4042 of the Inland Waterways Revenue Act of 1978). Included among specified waterways are the Mississippi River upstream from Baton Rouge, the Mississippi's tributaries, and the Gulf and Atlantic Intracoastal Waterways . The tax does not apply to fuel used by deep-draft ocean-going vessels, recreational vessels, or noncargo vessels such as passenger vessels and fishing boats . The tax rate of 8 cents per gallon was increased to 10 cents per gallon on October 1, 1985 [50] .

Certain aspects of the waterway user charge issue are relevant to any other alternatives . These include an assessment of Congressional intent in relation to the distinction among commercial, agricultural, flood control, power, and recreational use of the nation's waterways . Another aspect deals with the proper definition of the U .S. inland waterway system. Much discussion by policy makers has centered on the attempt to establish a user charge program that would be equitable for all sections of the nation [34] . The ultimate objective of a waterway user charge is to shift the burden of navigational improvements from general tax revenues, or the taxpayer in general, to the direct users of navigable waterways. The implicit consequence for the public would be a change of roles in the financial support for waterways [34] .

C. Rail Transportation Literature on rail transportation is summarized in the following subsections.

1. Domestic Railroad Network The railroad network of the U.S. consists of about 167,000 line miles. The operation of this

industry is distributed by 18 Class I railroads [41] . It includes some 481 ^{regional and local} railroads. All U.S. railroad carriers are legally classified as common carriers, which means that the interstate activities of all railroads are regulated by the federal government . These organization's intrastate activities are therefore subject to state regulation.

In 1929, railroads carried 74.9 percent of the total ton-miles of freight transported from one city to another. By 1979, the railroads transported only 36 percent of the intercity ton miles . Using the statistic of revenue ton-miles, nationwide the industry moved 858,105M revenue ton-miles of freight in 1978, which was 92 percent higher than the 447,332 carried in 1929 [70] . Revenue ton-miles are ton-miles of freight on which revenues are earned . They include most traffic carried. The example of non-revenue traffic would be a carload of railroad ties for the railroad's own use.

The railways of the United States are owned and operated by many different corporations. The Eastern District is owned and operated by Conrail, CSX Transportation, Florida East Coast, Grand Truck Western, Illinois Central and Norfolk Southern. The Western District ownership and operation is spread across Atchison Topeka & Santa Fe, Burlington Northern, Chicago & Northwestern, Denver & Rio Grande Western, Kansas City Southern, Soo Line, Southern Pacific and Union Pacific. All lines are connected and interconnected throughout the length and breadth of the country, forming one huge system. With a few minor exceptions, all the individual roads have the same gauge, and cars are permitted to move from one road to another . It is physically possible to convey a freight car between any two freight stations in the United States, however far apart they may be.

All railroads have all been built to accommodate prevailing currents of domestic trade . These currents or flows have developed because of the nature of the resources and industries of the different parts of the country . The internal trade of the United States, which the railroads were built to serve, has been based upon a sectional diversification of productive industry . The United States consists of three fairly well marked economic sections, a manufacturing and

commercial East, an agricultural South with cotton as its outstanding product, and an agricultural West, the staples of which have been grain and livestock . Railroad and motor carriers in the United States generally determine their rates,

classifications, and other charges through group consideration . Several economic characteristics aid in explaining the prevalence of rate making in the railroad industry . First, as common carriers they have public utility status and thereby are subject to federal regulation . Secondly, internal economies of scale, external economies of scale, joint cost, unused capacity, or some combination of these are common. These sizable elements of cost, known as supplementary costs, cannot be specifically identified with any particular traffic movement and consequently

must be allocated arbitrarily . Unused capacity on a round trip is usually the strongest cause of supplementary cost per unit of service [33].

In the railroad industry, the ratio of supplementary or fixed costs to variable costs has a propensity to be high because of heavy investments in railbeds, rolling stock, terminals, and other fixed facilities. An important dimension of the invariability of these costs is that costs per

unit of traffic tend to decline as traffic volume and plant utilization increases [31] . So long as under utilization or excess capacity exists, increases in aggregate traffic volume will reduce unit costs, as much as constant costs are allocated over a larger volume of tonnage . This condition,

in conjunction with an attendant inability to precisely identify common or joint movement costs,

creates a situation which eventually could result in destructive pricing in the absence of federal

or state rate regulation [41 ] .

2. Freight Traffic and Cost Structure

Freight traffic volume trends in the railroad industry may be examined in various ways, including car loading, tonnage originated, and ton-miles carried . Because ton-miles reflect both

weight and distance, they are normally considered the better indicator of actual service performed by a transportation mode compared to the other two ways . Revenue of freight in ton-miles generated by the railroad industry for 1990 was recorded at $1,034B.

The cost structure of the railroad industry is such that railroads can be very competitive in their short-term pricing policies and in the pricing of particular services [41] . Because of the very large investment in long-lived facilities such as track rights of way and terminals, a large portion of their costs are fixed or indirect in nature. During the life of these facilities, expenses of interest, depreciation, property taxation, maintenance, and similar costs do not vary with the amount of traffic handled. In the past, the cost structure has encouraged rate wars among railroads and between rails and other forms of carriage and the development of various kinds of discriminatory pricing policies [19] .

Revenues of freight in ton-miles, operating cost, expenses and income generated by different rail operation organizations are shown in Tables 2 to 6. The following are the Class I railroad companies reported on the Annual Report of Transport Statistics in the United States [24] :

• CR-Consolidated Rail

• CSX- CSX Transportation

• GTW - Grand Truck Western

• IC - Illinois Central

• NS - Norfolk Southern

• SL-Soo Line

• SP -Southern Pacific

• ATSF - Atchison Topeka & Santa Fe

• BN -Burlington Northern

• CNW- Chicago & North Western

• DRGW Denver & Rio Grande Western

• KCS - Kansas City Southern

• UP - Union Pacific

Table 2: Freight Revenues of Class I Railroads in The United States (thousands of dollars)

[24]

Table 3: Total Operating Revenues of Class I Railroads in The United States (thousands of dollars) [241 Table 5: Net Railway Operating Income of Class I Railroads in The United States (thousands of dollars) [241

1984 29,453,446 25,800,454 2,653,814 2,536,673 921,542 3,744,395 1985 27,586,441 25,225,295 1,788,151 1,746,386 876,984 4,422,903 1986 26,204,122 24,896,015 746,941 506,591 867,722 3,600,682 1987 26,6 22,482 23,878,116 1,965,475 1,756,460 943,747 2,970,805 1988 27,934,285 24,811,138 2,286,003 1,979,719 996,182 3,681,447 1989 27,955,969 25,037,666 2,009,094 1,894,315 1,013,821 3,708,662 1990' 28,369,803 24,651,542 1,961,127 2,648,258 1,033,969 3,639,838 1991r 27,845,206 28,061,187 (90,849) (37,455) 1,038,875 3,437,363 1992 28,348,741 25,316,364 2060,179 1,959,553 1,066,781 3,702,367

Table 6: Railway Revenue Ton-Miles and Expenses (millions of dollars) [241

Railroads have variable costs of operation . These are costs that vary directly with volumes . Examples of railroad variable costs are: maintenance of equipment and rights-of-way based on usage; labor costs; fuel, and lubrication oil.

Prior to the 1940's, many transportation experts believed railroad fixed costs were as much _as two-thirds of the total cost structure [41]. Today, it is generally believed that fixed costs are 40 to 50 percent of the total cost structure . The primary reason for this is that many railroads have greatly expanded their ton-mile production, using the same basic right-of-way and physical

plant. When volume increases over time, the percentage of fixed costs compared to total cost decreases .

Because of the still relatively high fixed cost structure of railroads, economies-of-scale apply with increasing freight. As volume increases, the total cost of production decreases on a per unit basis[7]. The reason is that asvolume increases, fixed costs stay constant and hence become less per unit of output. James C. Nelson [37], a noted railroad scholar, states : "Railroads, having a substantial investment in plant and equipment and relatively large fixed costs, cannot operate efficiently with low and irregular volumes of traffic . Either adequate traffic flows must be stimulated, or unprofitable operations and high rates will result ."

3. Cost Functions
According to Hasenkamp's model [16], the general term for a cost function is :

h(y) = g(x), where, (2) y = vector of output (consists of passenger-miles, freight ton-miles), x = vector of input (labor measured as man-hours, fuel measured as ton of coal &

capital services measured as car-miles or train hours) .

The analysis of cost function for railroad in this section is based on Translog Hedonic Cost Function [16] . The short-run variable cost function is C = C (F (y, q), x, v, t) ⁽³⁾

where,

C = a short-run variable cost.

F = measure of effective ton-miles .

y = (1 *N) vector of output .

q = (1 *R) vector of shipment attributes associated with each output .

x = (1 *H) vector of fixed factors.

v = (1 *J) vector of prices of the variable factor.

t = (1 *M) vector of technological variables .

N, R, H, J, M represent numbers of vectors in each different variable .

Studies show that this functional correlation has been inadequate in two respects . First, the

equation fails to distinguish the effect way-and-structures capital has on cost from that of track (route-miles). Second, the equation ignores the effect on costs of such variables astraffic mix by commodity type, average length of haul, and low-density route-miles [16] .

In specifying the rail cost function, Friedlaender & Spady [16] assume that way-andstructures capital such as track, bridge & building, signal, communication and others is fixed in the short run cost estimating calculation . This minimizes expenditures on the remaining factors given the exogenously determined quantity of the fixed factor, technological conditions, and outputs. It is appropriate to regard way-and-structure capital asfixed not only because its quantity is difficult to vary over short periods, but also because railroads are generally believed to have permitted this factor to decay rapidly in many instances . Since railroads are obligated by their status as common carriers to satisfy all shipper demands at fixed prices, both output and the technological conditions influenced by are exogenous [16]. Therefore, equation (3) has been modified as follows:

C = C (Yp, Yf; x, v, t) (3 a)

where,

yp = passenger service output, yf= freight service output,

v = vector of prices of variable factors,

x = quantity of way-and-structure capital,

t = vector of technological conditions .

Passenger service output is measured by passenger-miles, y p, which is then adjusted for another two quantities, passenger average travel length (PATL) and passenger miles per passenger route-mile (PDENS). Passenger service is defined as: yp= yp* f(PATL, PDENS), wheref isthe translog function. According to Friedlaender & Spady [16], one can expect that increases in passenger density will decrease the cost of producing a given amount of passenger-miles because of economies of utilization of equipment and centralization of ancillary service (ticket offices, terminal, etc .). Passenger service also requires that track be kept in better condition. Handling fewer passengers per passenger-mile lowers the cost associated with ticketing and boarding passengers . Increases in passenger average travel length are also associated with a shift in the type of passenger service provided from commuter to interstate . If the latter is more expensive than the former, we would expect dy p/dPATL < 0 for larger values of PATL.

For freight service output, revenue ton-miles is used . Since average length of haul and traffic mix are liable to have a direct effect on factor shares, they have been included as technological conditions, which is a less restrictive condition than including them in the Hedonic function altering ton-miles.

The following five variables are used in estimating rail expenditures : equipment, general and maintenance labor, traffic and transportation labor (other than train), on-train labor, and fuel

& material. According to Friedlaender & Spady [16], operating cost for railroad is equal to the sum of net way and structures, equipment depreciation, fringe benefit, labor taxes on employee compensation chargeable to operating expenses, net equipment rentals expenses, and imputed opportunity and depreciation cost on equipment capital.

D. Air Transportation Literature on air transportation is summarized in the following subsection .

1. Domestic Air Transportation System

Aviation became an exciting form of transportation during the past century . The domestic airline industry began in the mid and late 1920's. The government provided subsidies to the airline companies to carry mail . Mail transport was the beginning of the airline industry. Passenger business developed as an afterthought to the carriage of mail. Passenger aircraft technology burgeoned during the 1930's. In the late 1930's there were many developments which improved the dependability and safety of the air navigation system -an essential step for air travel to become popular. Not only have there been developments in aircraft, there have also been parallel developments in airports and air navigation equipment which have accommodated the increased size, sophistication, and dependability of aircraft .

U. S. scheduled commercial airlines now operate some 3,400 jet aircraft. The domestic revenue ton-miles of freight by the air industry in 1990 totaled $9.2B. Air transportation is still primarily a passenger operation, which generates over 80% of their revenues [39]. Air cargo is defined by the airline industry to include freight, express, and mail. Freight represents the biggest portion of the air cargo . Most air freight is transported by combination (both passenger and cargo) carriers either in the belly of scheduled passenger aircraft or in aircraft used

exclusively for cargo. Combination carriers usually find air freight service to be unprofitable, and all-cargo airlines have usually found it very difficult to be as profitable as common carriers of other modes [19] . Freight is often considered a byproduct of the passenger service in

combination carriers, because the aircraft will be flown with or without the freight. Some carriers believe that only those costs specifically attributable to carrying freight should be assigned to freight traffic for pricing purposes, rather than assigning to freight traffic a share of all the costs incurred [19] . Airline cost structures are more comparable to trucking industry cost structures than to railroad costs. This is because airlines have a relatively high variable costs in proportion to fixed costs [39] . Airports and airways navigational aids are provided through federal and states funds. This has made cost allocation a very difficult problem .

There are about200,000general aviation aircraft in theU.S .,compared with about3,000 airline aircraft [69] . General aviation aircraft provide flight service to communities without scheduled airline service [69] . At present, more than 700 airports receive scheduled airline service. There are approximately 15,000 airports (although many are unpaved and unlighted) . General aviation and commercial airlines compete with each other for air space and runway usage at major airports.

During 1989 there were 125 air traffic hubs in existence . They represented 25%of the 498 air traffic hubs and nonhubs in the 50 states, the District of Columbia, and other U.S. areas receiving air carrier service during the year . Air traffic hubs are not airports ; they are the ^cities or twin cities requiring aviation services. The dominance of the hubs in air traffic patterns is _{underscored by the fact that} 98% of passenger enplanements were recorded at these 125 hubs. Table 7 shows the number of hubs/nonhubs and the number of airports in those hubs . ^The

nonhubs are the communities enplaning less than 0.05 percent of the total passengers. The table also shows the number and percentage of passenger enplanements at the hubs/nonhubs .

Hub Numbers of Number of Enplaned Number Passengers Classification Hubs/Nonhubs Airports Percent

Large 28 53 313,779,281 73

Medium 34 39 76,097,543 18

Small 63 63 30,032,059 7

Nonhub 373 381 9,745,719 3

Total 498 544 429,654,602 100

Table 7: Numbers of hubs/nonhubs and the number of airports in those hubs [59]

In domestic aviation the aircraft are usually owned by private parties, either airlines or

general aviation aircraft operators. The airports to and from which they operate are typically

publicly owned, and the air navigation system they rely upon while airborne is provided by the

federal government.

2. Airports and Airways

Airports provide the interface between air and ground transport as well as fulfilling the _{usual terminal functions}. There are many types of airports, ranging from sophisticated airports found near major cities to small grass strips. Airports are classified by the Federal Aviation _{Administration} as (1) local interest airports, (2) national system of airports, and (3) military _airports [39]. Heliports, which can accommodate helicopters can be found on the tops of buildings in large cities. Seaplane bases service seaplanes and amphibious aircraft . These bases

are found in areas where conventional landing strips are unavailable . Airports generally are developed with public funds --federal, state, county, municipal, or a combination . Some airports are privately owned by either an industry or flight school . Airports used by commercial

airlines are provided by government. Charges are levied against airlines in the form of landing fees and rental and lease fees of various kinds for their use of the landing field, terminal building and other physical facilities.

Land requirements are great, especially off the ends of the runways, where approach paths must be kept clear of obstacles to aircraft flight or the pilot's vision . Early airports were developed with private funds . The larger passenger-handling airports were owned by either airlines or aircraft manufacturing companies. The situation changed in the mid-1930's when communities used federal public works programs for airport construction. The Federal Airport Act of1946 established a federal aid program which matched construction funds supplied by local and state interest [36]. In 1970, the federal program was increased, with part of the increased funding coming from user fees paid by both airlines and general aviation .

Airports are not used exclusively by airlines ; all airports used by the airlines are also used

.

extensively by private flyers of various sorts, and by flying services of the armed forces ^[30] Practically without exception, airport facilities have been provided by government expenditures, primarily by federal and municipal governments . The states have not been major players in the funding of airports. Since there are a variety of types of airport users, the problem of allocating proportionate costs for the use made of facilities becomes important ^{. Public airports are owned} and operated by city governments or special airport districts or boards . They attempt to recover

their cost from the users of the airports. Principal revenue sources for them are landing fees and space rental to airlines and concessionaires in and around the terminal building . Each passenger who uses the airport is charged in a number of ways to raise funds to meet the airport's costs . Most public airports budgets have two components : an operating costs budget and a capital cost

budget. Both usually cover a 12-month period. Operating costs include personnel costs, rent, supplies, and equipment with relatively short useful life. Capital costs are investments in land, buildings, and equipment with longer useful life.

Most air navigation facilities are provided by the federal government, operating through the

Federal Aviation Administration (FAA) . There are control towers at major airports ; traffic

control facilities controlling traffic above major metropolitan areas; and route navigation

devices, which emit signals used by cross country flyers to determine their location. Although

there are over 12,000 airports in the United States, only 600 have control towers. Only about

3,000 airports are both paved and lighted, meaning they are available for "round-the-clock"

operations every day of the year [69]. Airports are funded by combinations of federal, state,

county, municipal, and local money, and some of the costs are recovered through charges placed

on aircraft using the airport and passengers moving through the terminal.

Commercial airlines are classified into four categories by the Department of Transportation

[54]. The four carrier groups are:

• Major --Operating Revenues of $1,000,000,000 and above.

• Nationals -- Operating Revenues of $100,000,000 to $1,000,000,000 .

• Large Regional -- Operating Revenues of $10,000,000 to $99,999,999.

• Medium Regional -- Operating Revenues of $0 to $9,999,999 .

3. Cost Structure

The cost structure for air transportation is comprised of three major fixed costs which are route, terminal, and vehicle [19]. Some other general costs and variable costs also apply . Unlike surface carriers, air transportation does not have a physical route in the sense of a fixed physical roadbed or right of way . However, aircraft do make use of traffic control facilities designed to prevent accidents, and part of the system involves assigning air space to aircraft over a designed route, as well as control of traffic into and out of airports . This airway system, as it is called, is used by all kinds of aircraft (for hire, general aviation, and military) . The system is provided and operated by the federal government through the Federal Aviation Administration

(FAA).

Because of the lack of an investment in physical routes, the airlines avoid the associated necessary return to investors, amortization or depreciation, maintenance costs, and property taxes. Instead, the user charges paid by aircraft operators to help pay for the airway system are

in the nature of a variable cost because they tend to increase and decrease with the volume of traffic carried [39] . However, the route costs incurred by airlines are common in the sense that they are incurred in order to provide transportation of different passengers and/or different kinds of cargo traffic. Cost allocation to different segments is, therefore, difficult, particularly when the same aircraft carries both passenger and cargo traffic simultaneously

.

Airline passenger terminal facilities are places where passengers purchase tickets, transfer

between aircraft and airlines, and board and depart from aircraft ^{. Air cargo terminals are places}

where cargo is collected after local pickup and prior to line-haul by air . However, airports used

by commercial airlines are provided by government^{. Charges are levied against airlines in the}

to the large variable costs is the large amount of and high price of labor used in airline operations, a cost that tends to be variable. Air cargo has been expensive to handle because of the methods used, and the odd-shaped cargo (especially in combination carriers) .

Historically, the total operating costs of an airline have consisted of two major components : the direct operating costs and the indirect operating costs . The direct operating costs are those incurred as a necessary result of, and directly related to, flying the aircraft . The indirect operating costs are not directly related to the operation of the aircraft but are incurred as a result of operating services on the ground [45] . Direct operating costs of flying passengers or freight do not differ materially and, as a result, the elements and variations in direct costs for most operations are identifiable within the framework of the airline accounting structure [9]. The allocation of these costs in combination aircraft is still a matter of discussion and controversy .

4. Air Cargo Joint Cost Allocation.

When the airplane consumes fuel and oil to carry both passengers and cargo, this incurred cost is referred to asjoint cost[10]. Any attempt at allocating these costs is always based upon an individual value judgment, not upon an underlying, self-evident truth . Brewer and DeCoster

[9] indicated that all methods of allocating joint costs result in allocation either by physical attributes or the ability to carry costs. The allocation of costs by physical attributes is often done on the basis of the services rendered . Ton-miles, for example, implies that the costs incurred vary in direct proportion with variations in these physical attributes . ^{Also, all physical units are} treated as homogeneous in nature, which may not be true.

The allocation of costs to the services on the basis of their ability to absorb costs results in costs for each service that are proportional to the sales value^{. This can be considered primarily}

as a way of allocating the costs among the products [45]. Changes in the market value of one or more of the products automatically causes a change in the cost allocation basis. Both of these methods fall short of the desired results of truly accurate and reliable costing . Unfortunately, the air cargo industry has joint costs and there is no way of eliminating them .

5. Allocation ofAirway Costs.

A civil airway may be defined as a path through the navigable airspace ( above the minimum altitudes of flight prescribed by regulations) [30] . These airways are designated by naming points on the earth and connecting them with straight lines . The width of the airway extends five miles on either side of the center line .

If the cost of the airways is to be allocated among the various users, a decision must be made as to what costs should be covered and by whom . The determination of the portion of total costs to be met by the different users is part of this problem as is also the determination of the treatment to be given to historical costs or costs that have already occurred . Because of the nature of the reasons for expenditures upon airways, i.e., national defense requirements, rapid development of feeder airways, etc ., the federal government should not attempt to collect total allocated airways costs. Only current costs of the government in the maintenance and improvement of the airways should be considered [30] .

It would be inequitable for the airline passenger and shipper to be required to absorb a

portion of the cost of general aviation, since vast majority of such passengers and shippers could

not afford to own and operate an aircraft of their own.

The airways system is wholly owned and operated by the federal government. The cost of the federal airways system, and the allocation of that cost among users, can be determined on a

national basis. No fair determination can be made of airport costs on a national basis, and one

certainly could not allocate airport costs among users on a national basis . This must be done airport by airport. Any allocation of the airways system costs must be based on an appraisal of actual utilization of the airways corridor by each type of users . It is apparent that some users make greater use of the airways system than others . For example, on the basis of total number

of aircraft using the system, general aviation makes more use of the airways (98%) then either the airlines or military aviation. It can also be measured on the basis of flying hours . In terms of flying hours, the airlines have 18% of flying hours on the system [49] .

A 1966 FAA [49] report on "User Charges for the Domestic Federal Airways System", analyzed the extent of actual aircraft activity by type of operations involved under each of the thirteen major categories of airways facility groups. These includes air traffic control towers, approach control facilities, radar, instrument landing service systems, approach lighting,

VORTAC, air route traffic control centers, flight service stations, and others . This FAA analysis shows that different classes of users are dominant in the use of different airways facilities [49]

. The total aircraft operations on different categories of airways facilities are presented below:

FAA air traffic control towers

Air carrier 20.6% General aviation 70.3% Military 9.1

Flight service stations Air carrier 9.0% General aviation 73.5% Military 17.5%

Surveillance radar instrument Air carrier 62.9% General aviation 23.3% Military 13.8%

The over-all weighted average use of the airways facilities in the 1966 report [49]

shows the distribution of user percentage as

Air carrier 44.6% General aviation 29.3% Military 26.1

However, there is a shortcoming in the FAA report . It fails to assign sufficient weight to the paramount claim of the military on the airways system . The military use of the airways has some extraordinary characteristics . There are special services provided to military users of the airways by FAA. Among these special services are the "Restricted Airspace" for military activities, which are prohibited to civil aviation users, and the "priority" in the use of the airways

[49] .

All the government uses of the airways were classified under military uses . Many of them are either government users or non-public users, such as VIP flights and flights by the FAA, which owned over 100 aircraft. Many of these operations have been lumped, improperly, into

the general aviation category .

VI. Method of Analysis

A base cost model has been adopted from Hay [20] for modal cost assessment. Modification to this base cost model have been made, where required, for the different modes of

transportation that have been analyzed . The base cost model is described as : Total Transportation Cost = Capital Costs + Operating Costs (4)

where, Capital Costs = Facilities Costs + Equipment Costs (5) Operating Costs = Facilities Maintenance Costs + Equipment Maintenance Costs +

Transport Costs + Traffic Costs + General Costs (6) The definition of each cost component is described in greater detail below .

Capital Cost -Costs of providing initial plant and equipment and additions to or betterment of those facilities, such as initial roads construction, railroad tracks construction, ports construction,

etc. .

Operating Cost -
Costs of providing transportation services.

Facilities Cost -
Investment in routes, structures, and terminals .

Equipment Cost -
Investment in vehicles.

Facilities Maintenance Cost -
Costs of maintaining roadway and track, pavement and subgrade, rivers and harbors,
channels and dams, etc. .

Equipment Maintenance Cost -
Costs of maintaining motive power and rolling stock, such ^{as cars, locomotive, trucks,}
airplanes, tow boats, etc. .

Transport Cost -
Costs of conducting transportation, such as power and fuel, wages of vehicle crew, wages
of those directing vehicle movements .

Traffic Cost -

Costs of traffic solicitation, wages of highway safety officers, advertising, publishing

rates and tariffs, and administration .

General Cost -

Costs of general office expenses, legal advice, accounting, and salaries of general officers

and staffs.

A. Truck Transportation

There has been a great deal of research conducted over the years in determining and allocating costs of highway transportation . The method of cost allocation is based on several

cost allocation studies that have been conducted over the years.

The method divides highway transportation costs or expenditures into two kinds of costs. These two kinds of costs are attributable and non-attributable costs . Cost allocation methodology by the FHA [62] will be utilized for our analysis . The FHA analysis uses an improved incremental method of cost allocation for its analysis . Using the result of that analysis, all cost components in both capital costs and operating costs are carefully divided into attributable and non-attributable costs . Attributable costs are costs associated with vehicle

weight, while non-attributable costs are costs associated with vehicle mileage.

This cost allocation method uses the American Association of State Highway and

Transportation Officials (AASHTO) recommendation [47] . According to this organization, the

minimum thickness of a new pavement structure is considered a residual or non-attributable cost

shared by all vehicle classes. The extra thickness to accommodate traffic loading is assigned to

.

all vehicle classes by the AASHTO design parameter, equivalent single axle load (ESAL)

ESAL is a factor used by highway agencies to determine the overall pavement thickness required

_{for a new roadway}. This improved method eliminates the problem of the unfair assignment of the economies of scale in pavement cost/thickness that exists in the incremental method .

Vehicle Miles of Travel (VMT) was used in that study to determine non-attributable costs .

Truck transportation in our analysis considers the use of interstate highways only . This

simplifying assumption has been adopted because of the limited availability of cost data for

secondary, state, and local highways . The tables below show the shares of attributable and non-

attributable costs by each vehicle class. They are obtained from the study discussed above .

Attributable costs can be allocated to vehicles based on the percentage distribution presented in

Table 8 [47].

Vehicle Class Percent Share Autos/Motorcycles 21 .5 Buses 2.4 Pickups/Vans 9.1 Other single unit trucks 11 .9 Combination trucks 55.1

Table 8: Shares of Attributable Costs

Non-attributable costs can be allocated to vehicles in accordance with the percentage

distribution in Table 9 [47].

Vehicle Class Percent Share Autos/Motorcycles ^74.1 Buses 0.4 Pickups/Vans ^16.4 Other single unit trucks ^3.8 Combination trucks 5 .3

Three major cost components considered in our analysis of truck transportation are federal costs, state costs, and truck operating costs . The analysis, factors such as ratio of interstate truck mile to total highway system truck mile, ratio of number of bridges on interstate to total number of bridges on federal-aid highway system, and ratio of number of interstate vehicle miles to total number of federal-aid highway system vehicle miles are used to factor out cost responsibilities of truck on interstate .

1. Truck Transportation Cost Model

A set of modified cost models have been developed for the analysis of truck transportation . These cost models were derived from the base cost model (equations (4) - (6), pp.48) with the addition of several allocation factors. The cost models are defined as:

TTCt= CCt+ OCt (7) where,

CCt= FCt+ ECt (8)

OCt= FMCt+ EMCt+ TCt+ TRCt+ GCt (9) where,

FCt = a * b * (Federal Expenditures on Facilities + State Expenditures on Facilities) +

c * h * (Trucking Co . Expenditures on Facilities) (10) ECt= c * h * (Trucking Co. Expenditures on Equipment) (11) FMCt = a * b * (Federal Expenditures on Facilities Maintenance + State Expenditures on

Facilities Maintenance) +a * b * d * (Federal Expenditures on Bridge Maintenance) (12) EMCt = c * h * (Trucking Co . Expenditures on Equipment Maintenance) (13) TCt= c * h *(Trucking Co. Expenditures onTransport) (14) TRCt= e * b *f * (Federal Expenditures on Traffic) + *e * b * g *(State Expenditures on Traffic) (15)

GCt= e * b *f * (Federal Expenditures on General Costs) +e * b * g *(State Expenditures on General Costs) + c *h * (Trucking Co. Expenditures on General Costs) (16)

where,

a = % of attributable cost responsibilities for truck (Percent share of cost responsibility of single trucks and combination trucks from Table 8)

b = % of ton-miles carried by ICC regulated truck (Ton-miles carried by ICC regulated truck over the total ton-miles carried by ICC and Non-ICC regulated trucks)

c = % of interstate truck miles on total highway system
(Truck-miles traveled on interstate highways over total truck-miles traveled
on all highways)

d = % of number of interstate highway bridges on total federal-aid highway system (The number of bridges on interstate highways over the total number of bridges on federal-aid highways .)

e = % of non-attributable cost responsibilities for truck
(Percent share of cost responsibility of single trucks and combination trucks
from Table 9)

f = % of interstate vehicle miles on federal-aid highway system
(The number of vehicle-miles traveled on interstate highways over the total
vehicle-miles traveled on federal-aid highways .)

g = % of interstate vehicle miles on all road systems
(The number of vehicle-miles traveled on interstae highways over the total
number of vehicle-miles traveled on all road systems .)

h = % of trucking cost allocated per state
(The amount of public road mileage in each state over the total amount of
public raod mileage in the nation .)

Note: Allocation factor c is used to factor out the truck-miles traveled on interstate
highways from the total truck-miles traveled on total highway system .

Allocation factor g is used to factor out the vehicle-miles traveled on interstate highways from the total vehicle-miles traveled on all road system, including local roads.

where,

CCt (Capital Cost) -

Costs of providing initial plant and equipment and additions to or betterment of those

facilities in highway transportation, such as initial roads construction .

OCt (Operating Cost) -

Costs of providing transportation services in highway transportation .

FCt (Facilities Cost) -

Investment in routes, structures, and terminals in highway transportation.

EC, (Equipment Cost) -

Investment in vehicles, such as trucks in highway transportation.

FMCt (Facilities Maintenance Cost) -

Costs of maintaining roadway, pavement and subgrade in highway transportation .

EMCt (Equipment Maintenance Cost) -

Costs of maintaining motive power and rolling stock, such as trucks in highway

transportation.

TCt (Transport Cost) -

Costs of conducting transportation, such as power and fuel, wages of vehicle crew, wages

of those directing vehicle movements in highway transportation .

TRCt (Traffic Cost) -

Costs of traffic solicitation, wages of highway safety officers, advertising, publishing

rates and tariffs, and administration .

GCt (General Cost) -

Costs of general office expenses, legal advice, accounting, and salaries of general officers

and staffs in highway transportation .

Detailed descriptions of equations (10) to (16) are presented here . In equation ^(10),

allocation factor a is used to factor the attributable cost responsibility of trucks from

governments' expenditures (facilities costs) on interstate highways ^{. Since the analysis is only}

dealing with cost data from ICC regulated trucking companies, it is appropriate to factor the

portion of governments' expenditures (facilities costs) to ICC regulated trucks ^{. Allocation factor}

b is used for this purpose. Allocation factor c is used to factor the protion of trucking companies' expenditures (facilities costs) on interstate highways . Allocation factor h is used to factor the portion of trucking companies' expenditures on facilities costs for each state .

In equation (11), allocation factor c is used to factor the portion of trucking companies' expenditures on equipment on interstate highways . Allocation factor h is used to factor the portion of trucking companies' expenditures (equipment costs) for each state .

In equation (12), allocation factor a is used to factor the attributable cost responsibility of trucks from governments' expenditures (facilities maintenance costs) on interstate highways . Allocation factor b is used to factor the portion of governments' expenditures (facilities maintenance costs) to ICC regulated trucks. Allocation factor d is used to factor the portion fo governments' expenditures on interstate bridge maintenance.

The reason for the use of allocation factors in eqations (13) and (14) is the same as equation (11). In equation (15), allocation factor e is used to factor the portion of non-attributable costs of

trucks on governments' expenditures (traffic costs) on interstate highways . Allocation factor b is used to factor the portion of governments' expenditures (traffic costs) to ICC regulated trucks . Allocation factor f is used to factor the portion of federal government's expenditures (traffic costs) on interstate highways. Allocation factor g is used to factor the portion of state government's expenditures (traffic costs) on interstate highways .

The reason for the use of allocation factors in equation (16) is the same as equations (11) and (15) .

2. Truck Transportation Cost Data The tables on the following pages represent the expenditures by the federal government and

the respective state governments, and trucking companies . Information from Table 8 and Table 9 will be utilized to factor out the costs attributable to truck transportation . The following are the United States Department of Transportation's classification of highway expenditures . Specific costs by state are shown in Tables 11 to 20 [62] . The following

cost categories apply: Capital Outlay --Costs associated with highway improvements, including land acquisition and other right-of-way costs, preliminary and restoration costs of roadway and structure. Maintenance --Costs required to keep the highways in usable condition, such as routine patching repairs, bridge painting, and other maintenance of condition costs. Traffic Service --Costs associated with snow and ice removal, pavement markings,

signs, signals, litter cleaning, and toll collection expenses. Administration --Costs for general overhead and engineering, and research costs . Highway Safety --Highway law enforcement and safety expenditures. Bond Interest --Interest on the trust funds and bonds . Bond Retirement -- Retirement of trust funds and bonds. Intergovernmental --Grants-in-aid to local government.

Payments _{Interstate --}Costs associated with the development of interstate highway systems, including purchase of right-of-way and highway construction . _{Interstate --}Costs associated with highway repairs and pavement maintenance

. Resurfacing

Planning & --Costs for highway research and administrative planning .
Research

Bridge --Costs associated with replacing and restoring highway bridges. Replacement

Emergency Relief-- Costs associated with relief efforts during emergency period .

Others -- Costs associated with other categorial programs, beautification, interstate substitute, and other miscellaneous funds administered directly by FHWA.

Cost categories for equations (11) through (17) are described in Table 10 below .

Cost Components Cost Categories

Facilities Costs Interstate
Capital Outlays
Facilities Maintenance Costs Interstate Resurfacing
Maintenance

Traffic Costs Highway Safety
Emergency Relief
Traffic Service
Administration and Police

General Costs Planning & Research
Others
Bond Interest
Grant-in-aid
Bond Retirement

Table 10: Cost Components for Cost Equations

Federal and state highway expenditures are presented in Tables 11 through 20 . The data

presented are for the fiscal years of 1989-1991 .

In Tables 11, 13, 15, 17 and 19, the figures in all cost categories include the funds for other

road systems, such as primary and secondary roads, with the exception of interstate and

interstate resurfacing. The interstate portion of bridge replacement, will be factored out using

the ratio of the number of bridges on interstate system to the total number of bridges on federal

aid highway system. The interstate highway portion for rest of the cost categories will be determined by using the ratio of vehicle-miles of travel. The factor will be the number of interstate vehicle-miles over the total number of vehicle-miles on federal-aid system .

A different factor will be used in Tables 12, 14, 16, 18, and 20 to determine the interstate highway portion on each cost category . All the cost categories with the exception of capital outlays and maintenance, include funds for other road systems, such as freeways and arterial roads. The factor proposed will be the number of interstate vehicle-miles over the total vehicle-miles on all systems. Two of the three cases above will be using vehicle-miles traveled as the main factor, as this factor is a good representation of the traffic volume . This allocation factor was discussed in the cost-estimation techniques by Wohl [70].

Cost data for trucking companies were abstracted from the Blue Book [5] . The variable that will be used to factor out the interstate mileage will be the total number of interstate truck-miles traveled over the total number of truck-miles traveled on all systems . In order to calculate the cost for each state, the number of public road mileage for that state over the total number of

public road mileage for the nation will be used as the proration factor. The data on trucking

companies' expenditures is presented in Table 21 .

1989 1990 1991 Capital Outlays $198,349 $182,679 $151,505 Maintenance $23,573 $11,576 $12,719 Traffic Service $111,707 $104,800 $110,441 Admin. & Police $84,070 $98,573 $105,313 Bond Interest $13,829 $5,933 $4,945 Grant-in-aid $180,232 $149,810 $147,860 Bond Retirement $93,875 $21,415 $19,605

Table 12State Disbursements for Highway in Alabama [631 [641 [651

12
(thousands of dollars)

58

1989 1990 1991 Capital Outlays $49,749 $23,479 $28,017

Maintenance $10,205 $13,487 $12,542 Traffic Service $87,336 $83,365 $83,275

Admin. & Police $43,305 $48,666 $46,813 Bond Interest $--$--$-

Grant-in-aid $92,267 $85,467 $85,152 Bond Retirement $--$--$-

Table 14: State Disbursements for Highway in Arkansas [631 [641 [65] (thousands of dollars)

59

1989 1990 1991 Interstate $143,917 $152,050 $146,946

Table 1515: Expenditure of Federal Funds Administered by FHWA in Louisiana [63] [64]:

[65] (thousands of dollars)

1989 1990 1991

Capital Outlays $242,693 $201,340 $220,604
Maintenance $1,091 $2,338 $18,642

Traffic Service $47,367 $54,617 $42,602 Admin. & Police $139,953 $143,666 $135,266 Bond Interest $72,402 $62,892 $88,189 Grant-in-aid $47,863 $43,641 $32,845 Bond Retirement $340,396 $49,691 $66,390

Table 1616: State Disbursements for Highway in Louisiana [63] [64] [65]: (thousands of dollars)

60

[65] (thousands of dollars)

1989 1990 1991 Capital Outlays $42,454 $--$61,481 Maintenance $6,365 $--$5,369 Traffic Service $47,570 $51,396 $42,615 Admin. & Police $64,831 $59,529 $62,255 Bond Interest $16,772 $15,444 $15,100 Grant-in-aid $54,745 $57,905 $60,399 Bond Retirement $15,850 $17,796 $30,149

Table 18: State Disbursements for Highway in Mississippi [63] [64] [65] (thousands of dollars)

61

1989 1990 1991 Capital Outlays $154,342 $160,627 $95,691 Maintenance $44,149 $51,042 $48,140 Traffic Service $132,793 $153,489 $144,440 Admin. & Police $135,465 $132,271 $126,296 Bond Interest $2,055 $1,642 $1,237 Grant-in-aid $216,017 $28,538 ^$223,624 Bond Retirement $8,820 $8,820 $8,820

Table 20: State Disbursements for Highway in Tennessee [631 [641 [651

• (thousands of dollars)

All the costs from the tables are next used to obtain the necessary cost data for equations

(10) through(16). This yields a value for TTCtasspecified by equation(7). The determined TTCtrepresents the total transportation cost for the individual state.

The determined TTCt is then divided by the total ton-miles traveled per year by the trucking companies on the interstate highway system to obtain the true cost of transportation per ton-mile. In a mathematical representation, the true cost of transportation per ton-mile is, TTCt / total ton-miles traveled per year. Given the weight of the commodity and the distance traveled, one can determine the true effective cost to transport goods .

Cost data for trucking companies is available only from 1989 to 1991 . These data are used to determine TTCt and true cost per ton-mile for these three years . The true cost per ton-mile for the remaining years of 1992 and 1993 will be projected by using the least squares method suggested in Riggs [40]. The projected figures for the years 1992 and 1993 are summed with the figures for 1989, 1990 and 1991 . Mean values over the five year period are next determined . See Appendix A of Reference [10a] for detailed computations .

1989 1990 1991 Facilities Costs $2,699,065 $2,978,628 $3,262,338 Equipment Costs $8,293,777 $9,189,012 $11,127,110 Equip. Maint. Costs $1,871,146 $1,491,724 $2,128,070 Transport Costs $17,020,764 $16,844,705 $20,047,439 General Costs $13,504,473 $13,920,462 $16,595,170 Ton-Mile Carried 3.01 * 1011 3.11 * 10" 3.20 * 10¹¹

Table 21: Trucking Companies' Expenditures (thousands of dollars) [5]

3. Truck Transportation Cost Computation (A Numerical Example)

This section illustrate the truck cost calculation procedure . An example is used to show how the equations work, and how data and information is extracted from Tables 8 through 21 . As mentioned earlier, the calculation procedure is performed for each state for a given year . This example illustrates the analysis for the State of Alabama for 1989 .

The first step of this calculation procedure is to determine all the allocation factors for the State of Alabama for 1989 . As mentioned earlier, a is the allocation factor for the attributable cost for trucks. This is obtained by summing up the percent share of other single unit trucks (11.9%) and combination trucks (55.1%) in Table 8. This allocation factor remains the same for all three years in the analysis. The calculation forais as follows:

a = 0.119+0.551 = 0.67

Allocation factor b is defined as the percentage of ton-miles carried by ICC regulated trucks . This is obtained by dividing the ton-miles carried by ICC regulated trucks by the total ton-miles carried by ICC and non-ICC regulated trucks . These numbers are published in Transportation in America, by Eno Transportation Foundation [48]. This allocation factor changes every year . The calculation is as follows :

b = 302 billion ton-miles^-716 billion ton-miles = 0.422

Allocation factor c is defined as the percentage of interstate truck miles on total highway system. This is obtained by dividing the total truck miles traveled on interstate highway by the total truck miles traveled on all highways. These numbers are obtained from the publication, Highway Statistics [63] . This allocation factor also changes every year because the truck miles traveled changes every year. The calculation for this factor is as follows:

c = (67,628 million + 67,078 million) _ 605,456 million = 0 .223

Note: 67,628 million represents the truck miles traveled on urban interstate highways. 67,078 million represents the truck miles traveled on rural interstate highways .

Allocation factordis defined as the percentage of interstate highway bridges on the total federal-aid highway systems . This allocation factor changes every year and varies from state to state. This is obtained by dividing the number of bridges on interstate highways by the total number of bridges on the federal-aid highways. These numbers are also obtained from the publication, Highway Statistics [63] . The calculation for this factor is as follows :

d = (604 +554) _ 15437 = 0.075
Note: 604 represents the number of bridges on urban interstate highways .
554 represents the number of bridges on rural interstate highways .

Allocation factor e is defined as the percentage of non-attributable cost for trucks . This allocation factor is obtained by summing up the percent share of other single unit trucks (3 .8%) and combination trucks (5.3%) from Table 9. This factor remains unchanged over the three years of this analysis (same asthe allocation factor a). The calculation for this factor is as follows:

e = 0.038 + 0.053 = 0.091

Allocation factor f is defined as the percentage of interstate vehicle miles on federal-aid highway system. This allocation factor is obtained by dividing the vehicle miles traveled on interstate highways over the total vehicle miles traveled on federal-aid highway system . This allocation factor is to allocate federal expenditures to interstate highways . These numbers are

obtained from the publication, Highway Statistics [63] . This factor changes every year and varies from state to state. The calculation forf is as follows : f = 7,594million miles^-30,842million miles = 0.246 Allocation factor g is defined as the percentage of interstate vehicle miles of all road systems. This allocation factor is obtained by dividing the vehicle miles traveled on interstate

highways by the total vehicle miles traveled on all road system . This factor is to allocate some state expenditures to interstate highway. These numbers are obtained from the publication Highway Statistics [63] . This factor changes every year and varies from state to state . The calculation for g is as follows:

g = 7,594 millions miles _ 40,765 millions miles = 0.186

The difference between allocation factorfand allocation factor g is the denominator in both allocation factors. The denominator in allocation factor f is the total vehicle miles traveled on federal-aid highway system. The denominator in allocation factor g is the total vehicle miles traveled on all road system, which includes all federal-aid road systems and state funded road systems.

Allocation factor h is defined as the percentage of trucking companies' cost allocated per state. This allocation factor is obtained by dividing the total public road mileage in each state by the total public road mileage in the nation. These numbers are again obtained from the publicationHighway Statistics [63] . This allocation factor changes every year and varies from state to state. The calculation forh isasfollows:

h = 90,672 miles _ 3,880,151 miles = 0.0234

The data presented in Table 20 represent the figures for the entire nation's ICC regulated trucking companies. The second step of this calculation procedure is to apply these allocation factors to equations (7) through (16). The data used in this calculation is extracted from Table

11, 12, and 20. The following are the calculation examples for equations (7) through (16) . FCt= 0.67 * 0.422 * ($119,314,000 + $198,349,000) + 0.223 * 0.0234 * ($2,699,065,000) _ $104,325,970

$119,314,000 is federal expenditures on facilities on Alabama's interstate highways in 1989, and $198,349,000 is state expenditures on facilities on Alabama's highway systems in 1989 . Both these numbers are extracted from Table 11 and Table 12 according to the classification in Table 10. $2,699,065,000 is the trucking companies' expenditures on facilities in 1989, and this number is extracted from Table 21 .

ECt= 0.223 * 0.0234 * ($8,293,777,000) = $43,278,590 $8,293,777,000 is the trucking companies' expenditures on equipment in 1989 . This number is extracted from Table 21 . FMCt= 0.67 * 0.422 * ($56,676,000 + $23,573,000) + 0 .67 * 0.422 * 0.075 ($49,516,000) = $23,852,130 $56,676,000 is the federal expenditures on facilities maintenance on Alabama's interstate highway in 1989, and $23,573,000 is state expenditures on facilities maintenance on Alabama's highway systems in 1989. Both these numbers are extracted from Table 11 and Table 12 according to the classification in Table 10 . $49,516,000 is the federal expenditures on bridge maintenance on Alabama's interstate highway in 1989 . This number is extracted from Table 11 . EMCt= 0.223 * 0.0234 * ($1,871,146,000) = $9,764,010

$1,871,146,000 is the trucking companies' expenditures on equipment maintenance in 1989. This number is extracted from Table 21 . TCt= 0.223 * 0.0234 * ($17,020,764,000) = $88,817,750 $17,020,764,000 is the trucking companies' expenditures on transport costs in 1989 . This number is extracted from Table 21 . TRCt= 0.091 * 0.424 * 0.246 * ($6,802,000) + 0.091 * 0.424 * 0.186 * ($111,707,000 + $84,707,000) _ $1,469,570

$6,802,000 is federal expenditures on traffic on Alabama's highway systems in 1989 . This number is extracted from Table 11 according to classification in Table 10 . $111,707,000 and $84,707,000 are state expenditures on traffic on Alabama's highway systems in 1989. Both

these numbers are extracted from Table 12 according to classification in Table 10 . GCt= 0.091 * 0.424* 0.246 * ($4,120,000 + $751,000) + 0.091 * 0.424* 0.186 ($13,829,000 + $180,232,000 + $93,875,000) + 0 .223 * 0.0234 * ($13,504,473,000) = $72,579,770

$4,120,000 and $751,000 are federal expenditures on general costs on Alabama's highway systems in 1989. These numbers are extracted from Table 11 according to classification in Table 10. $13,829,000, $180,232,000 and $93,875,000 are state expenditures on general costs on Alabama's highway systems in 1989 . All these numbers are extracted from Table 12 according to classification in Table 10. $13,504,473,000 is the trucking companies' expenditures on general costs in 1989. This number is extracted from Table 21 .

TTCt= CCt+ OCt = FCt+ ECt+ FMCt+ EMCt+ Tct+ TRCt+ GCt = $104,324,970 + $43,278,590 + $23,852,130 + $9,764,010 + $88,817,750 + $1,469,570 + $72,579,770

= $334,086,790

The third step of the calculation is to determine the portion of ton-miles carried for Alabama in 1989. The data used in this calculation is obtained from Table 21. Below is the procedure and equation for this calculation.

Ton-Mile Carried for each state = c * h * Ton-Mile Carried (17) Ton-Mile Carried for Alabama in 1989 = 0.223 * 0.0234 * 3.01 * 10¹¹ = 1,570,678,200 Cost per ton-mile = TTCt - Ton-Mile Carried = $334,086,790--1,570,678,200 = $0.213

3.01 * 1011 is the ton-miles of freights carried by ICC trucking companies in 1989. This number is extracted from Table 21.

B. Barge Transportation

There is probably no field of transportation that has a greater dearth of data and studies of economic analysis than barge transportation . In order for a waterway to be navigable, facilities such as canals, lock and dams, and ports are needed. The funds to develop these facilities come from both the federal and state government and also from private businesses . The functions of waterways and canals are not solely for freight navigation purposes, they are also used for flood

control, drinking water, power generation, and recreation .

1. Barge Transportation Cost Model

The cost equations for barge transportation are :
TTCb = CCb+ OCb
CCb = FCb+ ECb

OCb = FMCb+ EMCb+ TCb+ TRCb+ GCb ₍₂₀₎

where,

CCb (Capital Cost) -Costs of providing initial plant and equipment and additions to or betterment of those facilities in waterway transportation, such as ports construction and canal construction .

OCb (Operating Cost) -
Costs of providing transportation services in waterway transportation .

FCb (Facilities Cost) -
Investment in routes, structures, and terminals in waterway transportation.

ECb (Equipment Cost) -
Investment in vehicles in waterway transportation, such as tow boats and barges .

FMCb (Facilities Maintenance Cost) -
Costs of maintaining rivers and harbors, channels and dams, etc . .

EMCb (Equipment Maintenance Cost) -
Costs of maintaining motive power and rolling stock, such as tow boats.

TCb (Transport Cost) -Costs of conducting transportation, such as power and fuel, wages of vehicle crew, wages of those directing vehicle movements in waterway transportation.

TRCb (Traffic Cost) -Costs of traffic solicitation, advertising, publishing rates and tariffs, and administration in waterway transportation.

GCb (General Cost) -Costs of general office expenses, legal advice, accounting, and salaries of general officers and staffs in waterway transportation .

The focus of this analysis is on cost allocation on multiple purpose projects for rivers and

other inland waterways. The cost distribution guideline defined in the Water Resources

Development Act of 1985 [50] will be used to prorate these projects . In the guideline, 35% of

costs on irrigation/navigation projects are allocated to navigation. The same cost allocation factor applies to flood control/navigation projects . The cost distribution for recreation/navigation projects is 50/50. Under the same guideline, 100% of the costs for hydroelectric and water supply projects are solely for their own purposes .

2. Barge Transportation Cost Data

Cost information for the categories previously defined has been extremely difficult to extract from the literature, to date . At this point, we are working closely with the Arkansas Waterways Commission to identify other sources of cost information that have been published by the Army Corps of Engineers.

Cost information for the categories previously defined has proven to be extremely difficult to extract from the literature. All the cost information used in this analysis are provided by the Institute of Water Resources, Water Resources Support Center of The U.S. Army Corps of Engineers [14,15] .

The resultant cost model for barge transportation is similar to the other cost models. As mentioned earlier, the allocation factors separate the commercial navigation users' cost responsibilities from other users . This task was found to be unnecessary as the cost data provided by the Institute of Water Resources [14] on the waterways had been separated by user category. The data used is solely for navigation purposes. The data provided by the Institute of Water Resources is only for the fiscal year 1994.

Table 22 shows all the cost data used in this analysis to determine the true cost of water freight transportation. Note that in Table 22, Facilities Costs and Facilities Maintenance Costs are combined into one cost category. These costs are the government contributions to the waterway infrastructures. The data in Table 22 is for 1994 .

3. Barge Transportation Cost Computation (A Numerical Example)

This section illustrates the waterway cost calculation procedure . An example is used to show how the equations work, and how cost information is extracted from Table 22 . As mentioned earlier, the data provided by the Institute of Water Resources are data compiled for the entire Mississippi River and its tributaries, and Gulf Intracoastal Waterways. It is virtually impossible to segregate them into per state basis. Therefore, this analysis calculates the cost on the system wide basis .

No allocation factor is required in this calculation as the cost data provided is solely attributable to navigation. The first step in this calculation is to extract the cost data from Table 22 and insert the data into equations (18) through (20). The calculation for equations (18) through (20) are presented as follows:

TTCb = CCb+ OCb = FCb+ ECb+ FMCb+ EMCb+ TCb+ TRCb+ GCb = $152,937.4 + $1,379,755.92 + $305,468.7 + $1,438,004.2 + $345,117.58

72

= $3,621,283 .8 * 103 The second step of the calculation is to determine the cost per ton-mile for water transportation. The data used in this step came from Table 22 and the previous calculation. Ton-Miles Carried = 239 * 109 (from last row in Table 22)

Cost per ton-mile = TTCb - Ton-Mile Carried
= $3,621,283,800 = 239 * 109
= $0.015

C. Rail Transportation

The construction of maintenance of railroad transportation routes are carried out solely by the respective operators . Costs involved in the operation include roadbeds, bridges, crossing, right of ways, etc.. There are no government contributions to the railway infrastructures for freight services. The federal government's contribution to railroads is mainly for Amtrak and Northeast Corridor Project. Since our analysis addresses costs only for the transportation of freight, it can be assumed that the total cost incurred by the operators in providing the service has a direct indication of the true cost of transportation in rail services . Therefore, the method of analysis is the most straightforward of the four transportation modes being analyzed .

1. Rail Transportation Cost Model
The cost model for rail services derived from the base cost model is defined as :
TTCr=CCr+ OCr (21)

where, CCr= a * (FCr+ EC) (22)

OCr= a * (FMCr+EMCr+TCr+TRCr+GCr) (23) where, a = % of track miles per state where, CCr (Capital Cost) -

Costs of providing initial plant and equipment and additions to or betterment of those facilities in rail transportation, such as railroad tracks construction. OCr (Operating Cost) -

Costs of providing transportation services in rail transportation.

FCr (Facilities Cost) -
Investment in routes, structures, and terminals in rail transportation .
ECr (Equipment Cost) -

Investment in vehicles in rail transportation, such as locomotives, box cars, etc .

FMCr (Facilities Maintenance Cost) -
Costs of maintaining railroad tracks, etc .
EMCr(Equipment Maintenance Cost) -

Costs of maintaining motive power and rolling stock, such as locomotive and box cars . TCr (Transport Cost) -

Costs of conducting transportation, such as power and fuel, wages of vehicle crew, wages
of those directing vehicle movements in rail transportation.
TRCr (Traffic Cost) -

Costs of traffic solicitation, advertising, publishing rates and tariffs, and administration in

rail transportation.
GCr(General Cost) -
Costs of general office expenses, legal advice, accounting, and salaries of general officers
and staffs in rail transportation .

The only cost allocation question in our analysis is the share of passenger service compared to the share of freight service. The unit of measurement in the railway cost allocation analysis is

the amount of cost attributable to each type of service. Traditional units like ton-miles or passenger-miles are not an appropriate allocation factors . Both units reflect only the attributes of

one type of service. Ton-mile reflects the freight service while passenger-mile reflects the passenger service.

2. Rail Transportation Cost Data and Allocation Procedure

Tables 23 shows the data used in the analysis of rail transportation cost . The source of the data is the annual report of the Interstate Commerce Commission [24] . The annual report shows both freight expenditures and passenger expenditures. Therefore, it is only necessary to include the freight expenditures in this analysis. Each cost category in the railroad companies' operating data is fitted into the base cost model to determine the actual cost of providing rail services . Mean values of a five year period are again determined. See Appendix A of Reference [10a] for detailed computations. Table 23 below represents the data used in the calculations of true transportation cost for railways .

1989 1990 1991 Total Passenger Expenditures $125,735,000 $127,328,000 $140,105,000 Total Freight Expenditures $24,911,931,000 $24,524,214,000 $27,921,082,000 Total Operating Expenditures $25,037,666,000 $24,651,542,000 $28,061,187,000 Total Ton-Miles Carried 2,128,275,904,000 2,185,006,708,000 2,150,235,017,000 Total Track Miles Operated 138,006 133,189 129,672

Table 23: Railway Expenditures and Traffic Data [24,25,26]

3. Rail Transportation Cost Computation (A Numerical Example)

An illustration of the railway cost calculation procedure is presented in this section. It shows how data and information is extracted from Table 23, and how the equations work . As mentioned earlier, the method of analysis for railway is the most straightforward of the four

transportation modes being analyzed . The illustration below demonstrates the procedure for the analysis for the State of Arkansas in 1989 .

The first step of this calculation is to determine all the allocation factors for the State of Arkansas in 1989. Allocation factor a is defined as the percentage of track miles per state . This is obtained by dividing the number of track miles in each state by the total number of track miles in the nation. The number of track miles for each state in this analysis is obtained from Maplnfo, the mapping software used in this project. Table 24 shows the track miles operated in each state. The calculation for a is as follows:

a = 3,195.48 miles^-138,006 miles = 0.0232

The next step in the calculation procedure is to apply the allocation factor to equations (21) through (23). All cost components in the equations have the same allocation factors . Therefore, they are combined into equation (21). The calculation for total transportation cost for railroads is as follows:

TTCr = a * (FCr+EC, + FMCr+EMcr+TCr+TRCr+ GCr)

= 0.0232 * ($24,911,931,000)

= $576,826,929.8 $24,911,931,000 represents the total freight expenditures incurred by railroad companies in 1989. This number is extracted from Table 23 under total freight expenditures .

The third step in this calculation is to determine the portion of ton-miles carried for Arkansas in 1989. The data used in this calculation is obtained from Table 23 . Below is the procedure and equation for this calculation.

Ton-Miles Carried for each state = a * Ton-Miles Carried
Ton-Miles Carried for Arkansas in 1989 = 0.0232 * 2,128,275,904,000 ton-miles
= 49,376,000,970 ton-miles

Cost per ton-mile = TTCt ^-Ton-Miles Carried

= $576,826,929.8 - 49,376,000,970 ton-miles

= $0.0117 per ton-miles

Note that in this analysis, allocation factor a is used in both numerator and denominator of cost per ton-mile calculation. Both allocation factors will canceled out each other. Therefore, cost per ton-miles can be calculated directly from dividing the total freight expenditures by the total ton-miles carried. Results from the lack of information on ton-miles of freight carried by each state is every state will have the same cost per ton-miles . Cost per ton-mile by state will be accurately reflected only if we have more detailed data on ton-miles carried per state.

D. Air Transportation

The cost responsibilities of air freight are distributed to all cargo carriers and combination carriers. This procedure presented in this section determines the portions of cost to be attributed to. all cargo carriers and combination carriers. In 1966, Civil Aeronautics Board (CAB) implemented Form 242 in an attempt to determine cost of all-cargo operations for combination and all-cargo carriers [11]. The cost formulas from this form were applied by the Civil Aeronautics Board (CAB) to the airlines' annual reports to the Board . All carriers are required to submit their operating expenses and other financial statistics in accordance with the Board's Uniform System of Accounts and Reports (Form 41) .

1. Cost Allocation Procedure

In reference to the cost formulas of Form 242, we have applied the following allocation

factors to the cost items reported in the Air Carrier Financial Statistics [54,55] . The allocation

factor was generated from the information and data provided in the Air Carrier Traffic Statistics

[56,57] .

Cost Items Allocation Factor Aircraft & Traffic Servicing Transport Related # of revenue ton-miles of freight Promotion & Sales # of total revenue ton-miles General & Administration Depreciation and Amortization Maintenance Flying Operation

Table 25: Allocation Factor for Air Carriers

Only the primary and commercial service airport reported in National Plan of Integrated Airport

Systems [67] are considered in our analysis. The Air Transport Association's (ATA) adjusted relative percentage of use by the different groups is used in this analysis instead of the finding from the 1966 FAA report [49] . The comparative percentages from each report are compared

below.

The reason for not choosing the FAA report is because it fails to assign sufficient weight to the paramount claim of the military on the airways system . The military use of the airways has some extraordinary characteristics. There are special services provided to military users of the airways by FAA. Among these special services are the "Restricted Airspace" for military activities, which are prohibited to civil aviation users, and the "priority" in the use of the airways. All the government uses of the airways were classified under military uses . Many of them are either government users or non-public users, such as VIP flights and flights by the FAA, which owned over 100 aircraft. Many of these operations have been lumped, improperly, into the general aviation category.

2. Air Transportation Cost Model
The modified cost model for air transportation is defined as :
TTCa= CCa+ OCa (24)

where,

CCa= FCa+ ECa (25)

OCa= FMCa+ EMCa+ TCa+ TRCa+ GCa (26)

where, FCa= a * b * c * (Government Expenditures) (27) ECa= a *d * (Air Carrier Expenditures) (28) FMCa= embedded in FC EMCa=a *d *(Air Carrier Expenditures) (29) TCa = a * d * (Air Carrier Expenditures) (30) TRCa=a *d *(Air Carrier Expenditures) (31)

GCa = a * d * (Air Carrier Expenditures) ₍₃₂₎

where,

a = % offreight revenue ton-miles over total revenue ton-miles
(Amount of freight revenue ton-miles carried over the total amount of
revenue ton-miles carried.)

b = % ofaircraft activity per state (airport traffic control towers)
(Amountofflights in each state over the total amountofflights in the
nation.)

c = % ofair carrier traffic at FAA facilities (airport traffic control towers)
(Amount of air carrier flights over the total amount of flights in the nation .)

d = % offreight activity per state
(Amountoffreight tonnage enplaned in each state over the total amountof
freight tonnage enplaned in the nation.)

where,

CCa (Capital Cost) --Costsofproviding initial plant and equipment and additions to or betterment ofthose facilities, such as airports, airways, aircraft, etc .

OCa(Operating Cost) --
Costsofproviding transportation services.

FCa (Facilities Cost) -
Investment in routes, structures, and terminals in air transportation.

ECa (Equipment Cost) -
Investment in vehicles in air transportation .

FMC,, (Facilities Maintenance Cost) --
Costsofmaintaining runways and lighting in airports, airways, etc.

EMCa(Equipment Maintenance Cost) --
Costsofmaintaining motive power and rolling stock, such as airplanes in air transportation .

TCa (Transport Cost) --
Costsofconducting transportation, such as power and fuel, wagesofvehicle crew, wages
ofthose directing vehicle movements in air transportation .

TRCa (Traffic Cost) -
Costs of traffic solicitation, advertising, and publishing rates and tariffs in air transportation .

GCa (General Cost) -

Costs of general office expenses, legal advice, accounting, and salaries of general officers

and staffs in air transportation.

The allocation factor to determine the portion of government expenditures on airports and airways per state is expressed as the ratio of the amount of air traffic activity for each state over the total amount of air traffic activity in the nation . The allocation factor to determine the portion of commercial activity in airports and airways is the ratio of number of commercial air carrier traffic activity at FAA facilities in a given state over the total number of air traffic activity at FAA facilities . It is realized that this allocation factor may not accurately reflect the actual cost responsibility of commercial carriers on airport costs. However, we feel that this is the best approximation possible due to the difficulty in obtaining more detailed cost data and information from the existing publications .

The allocation factor for determining the freight cost from the total commercial carriers costs is the ratio of number of freight revenue ton-miles enplaned over the total revenue ton-miles enplaned. Percentage of tonnage of freight enplaned in each state over the total tonnage of freight enplaned in the nation is the allocation factor used to prorate each state's portion of air carrier expenditures and ton-miles traveled.

As a general rule in this analysis, all the data will be collected for a five year period . In the event where data is not available for all five years, data from the most recent three years is collected. A linear regression model is used to determine the missing figures for the remaining two years. The true cost of air transportation is determined by averaging the annual TTC a per

^{ton-miles over the five year period}. The annual TTCaper ton-mile will be calculated by the diving the annual TTCaby the annual ton-miles of freight carried. See Appendix A of Reference [10a] for detailed computations .

3. Air Transportation Cost Data Table 26 shows the data used in analyzing and estimating the air transportation cost . The

data sources are Air Carrier Financial Statistics [54,55] and Air Carrier Traffic Statistics [56,57].

4. Air Transportation Cost Computation (A Numerical Example)

Following is an example to illustrate how equations (24) through (32) work and how the data in Table 26 were extracted for the analysis. The calculation procedure here is similar to analyses for other transportation modes. It is performed for each state and for individual year. This example illustrates the analysis for the State of Tennessee for 1989 .

The first step of this calculation procedure is to determine all the allocation factors for the State of Tennessee for the year 1989. As mentioned earlier, a is defined as the percentage of

freight revenue ton-miles carried . This is obtained by dividing the freight revenue ton-miles carried by the total revenue ton-miles carried shown on Table 26. This allocation factor is the same for all the states but varies from year to year . See Appendix C of Reference [10a] for allocation factors for 1990 and 1991. The calculation for a is as follows : a = 7,526,736 revenue ton-miles _ 42,456,749 revenue ton-miles = 0.177

Allocation factor b is defined as the percentage of air traffic activity per state. This is obtained by dividing the amount of air traffic in each state by the total amount of air traffic in the nation. These numbers are obtained from theFAA StatisticalHandBookofAviation[60]. This factors vary every year and from state to state . See Appendix C of Reference [10a] for allocation factors for1990and1991. The calculation forbisasfollows:

b = 961,716 flights - 61,345,173 flights= 0.016

The allocation factor c is defined as the percentage of air carrier traffic at FAA facilities. This is obtained by dividing the amount of air carrier traffic by the total amount of aircraft traffic. These values are obtained from theFAA StatisticalHandBookof Aviation[60]. This factor varies only from year to year. See Appendix C of Reference [10a] for allocation factors for 1990 and 1991. The calculation for c is as follows;

c = 20,816,616 flights^-61,299,017flights = 0.3393

The allocation factor d is defined as the percentage of freight activity for each state . This is obtained by dividing the freight tonnage enplaned for each state by the total freight tonnage enplaned in the nation. These parameters are obtained from theFAA StatisticalHandBook of

Aviation [60]. This factor varies annually and from state to state . See Appendix C of Reference [I Oa] for allocation factors for 1990 and 1991 . The calculation for d is as follows: d = 601703.12 tons- 3967009.7 tons = 0.1517

The next step is to calculate the true cost of transportation using equations (27) through (32) by applying these allocation factors with data extracted from Table 26 . Note that equations (28) through (32) have the same allocation factors and cost components. Therefore they are combined with equation (27) into equation (24). Following are the examples of the calculation for equation (24).

TTCa= 0.177 * 0.016 * 0 .3393 * ($11,017,000,000) + 0.177 * 0.1517 * ($52,260,462,000) _ $1,413,826,648 The next procedure in the calculation is to determine the ton-miles carried in the State of Tennessee in 1989. The data used in this calculation is obtained from Table 26 . Ton-Miles Carried for each state = d* Ton-Miles Carried (33) Ton-Miles Carried for Tennessee in 1989 = 0 .1517 * 7,526,736,000 ton-miles = 1,141,805,851 ton-miles Cost per Ton-Miles = TTCa _ Ton-Miles Carried _ $1,413,826,648 ^-1,141,805,851 _ $1.238 per ton-miles

VII.

Results of Analysis

A.

Data Collection An interim report on this project was submitted to the grantor in the Spring of 1994 . Since then, data collection has been somewhat successful, considering the limited of number of sources available. All data needed were gathered by the investigators for a complete analysis .

For highway/truck transportation, the data collection effort is complete . The data include the expenditures by federal and state government on interstate highway system, and operating data from ICC regulated trucking companies for the period 1989 to 1991 . Data collection for barge transportation has proven to be a very formidable task . A great deal of effort has been expended in trying to aggregate government expenditures on inland waterways system . There are no federal regulatory agency such as the ICC for truck and rail transportation . Data for

waterway analysis were gathered from the Institute of Water Resources .

The data collection for rail transportation has focused only on ICC Class I railroad company. Data has been aggregated for twelve Class I railroad companies over a period from 1989 to 1991 . This data has been published by ICC on its annual report [24] . Operating data for commercial air carriers was compiled from literature published by the Department of Transportation's Research and Special Programs Administration [54] . These carriers include all-cargo carriers and combination carriers from the carrier groups classified as majors, nationals, large regional, and medium regional. The applicable period is from 1989 to 1991. Annual government expenditures on the entire national airways are available . Unfortunately, this information is not tabulated on a state-by-state basis. Government allocation for expenditures on airports are available on airport by airport basis .

A preliminary version of transportation cost analysis software is available . An interim version this software will allow the user to estimate and input the required cost parameters where

data is not yet available. The final version of the software will contain a complete data for all transportation modes.

B. True Transportation Costs Generated from Analysis

The true transportation costs in terms of $ per ton-mile for truck, barge, rail, and air transportation are summarized in this section .

1. Truck Transportation

The true (societal) cost of providing truck transportation service determined from the developed methodology is presented in Table 27. Values presented are averages for the five states obtained from the data in Appendix A of Reference [1Oa] .

State True Cost of Transportation Alabama $0.215 per ton-mile Arkansas $0.177 per ton-mile Louisiana $0.278 per ton-mile

Mississippi $0.183 per ton-mile Tennessee $0.211 per ton-mile

Table 27: Societal Costs of Truck Transportation by State

2. Barge Transportation

Barge transportation costs are indicated for all states and are summarized below .

State True Cost of Transportation Alabama $0.015 per ton-miles Arkansas $0.015per ton-miles Louisiana $0.015 per ton-miles

Mississippi $0.015 per ton-miles

Tennessee $0.015 per ton-miles

Table 28: Societal Costs of Barge Transportation by State

3. Rail Transportation

The true cost of providing rail transportation service in terms of cost per ton-mile using our

methodology are summarized in Table 29. Values are averages for all states as previously

described.

State True Cost of Transportation Alabama $0.0115 per ton-miles Arkansas $0.0115 per ton-miles Louisiana $0.0115 per ton-miles

Mississippi $0.0115 per ton-miles

Tennessee $0.0115 per ton-miles

Table 29: Societal Costs of Rail Transportation by State

4. Air Transportation

The true cost air freight transportation in terms of cost per ton-miles determined by the

analysis are presented in Table 30. Entries shown are averages over the five states obtained

from the data in Appendix A of Reference [IOa].

State True Cost of Transportation Alabama $1.813 per ton-miles Arkansas $4.812 per ton-miles Louisiana $1.785 per ton-miles

Mississippi $6.712 per ton-miles

Tennessee $1.367 per ton-miles

Table 30: Societal Costs of Truck Transportation by State

VIII.

Development of Transportation Cost Software

A.

Introduction The objective of this section is to present a detailed description of the development of an interactive cost analysis software for calculating the time cost for truck, barge, and air transportation.

The software has been written in Windows environment . This is because windows application is very widely used and easily accessible . Pop-up windows have been generated to prompt the user to enter input data associated with the analysis region to determine the true transportation cost and expected delivery time . Pop-up windows have also been used to describe the cost assessment methodology to the user . A visual geographic display is used to interface with the developed software. A commercial software package called Maplnfo has been selected for this purpose. This will enable the user to visualize the display of the transportation networks

and the output of the analysis.

The software is efficient and makes use of advanced graphics, charts and maps . All pop-up windows are generated before the user enters input, so that the required information can be correctly entered for accurate analysis .

The software serves as a fundamental tool for the user to determine both the published and societal costs of every mode of the transportation. The ultimate goal of this software is to offer the user the option of combining and evaluating the different combinations of transportation modes in determining the minimum cost transportation alternatives. The software is being designed so as to be expandable to other analysis regions .

B. Visual Display Software

In order to provide visual displays of the region that is being analyzed, visual display software is needed. Maplnfo has been selected to provide the display .

1. MapInfo

MapInfo is known to be a powerful analytical resource which offers unique ease of use . The software offers a variety of maps, data including streets, railroads, waterways, and airport maps for the entire U.S. and part of Canada. Its data library has a program called StreetInfoTM, a

digital map data, which includes streets and street names, highway and highway names, water boundaries, for the U. S. These enable the user to see the selected region and its available transportation routes. Maplnfo offers a file called Airports-Basic, which contains information on over 17,000 airports in the U.S. and outlying territories. Maplnfo also includes data files containing the facility name, type of facilities (airport, heliport, seaplane port, etc .) and field coordinates. Railroad-Map is a file contained in MapInfo which includes U.S. railroads for the 48 contiguous states.

There is a technical software package called American Digital Cartography (ADC) WorldMap available for MapInfo. ADC WorldMap is a digital chart of the entire world and is considered to be the most detailed set of world-wide data available as well as the most comprehensive seamless set of data for North America . There are 17 layers of information and a MapBasic application which assists the user in loading the information . These layers include roads, railroads, utilities (pipelines, power lines), airports, rivers and lakes, bridges, tunnels, etc . MapInfo allows the user to zoom-in to see the areas of interest and also allows an attachment of a user data to any object on a map. Maplnfo's drawing tools enable the user to draw delivery

^{routes, defined regions, etc}. Maplnfo can directly read and query on dBASE, delimited ASCII, Lotus, and Excel files. Maplnfo has other tools that might be very useful to the user. These

include:

MapCode -A programming language for DOS .

MapBasic -A programming language for windows, Macintosh, Sun and HP.

Real Time Maplnfo -a direct serial communication with DOS applications to track many

objects and monitor changing status at remote sites .

Many Maplnfo users have concluded that this software is a very powerful and useful tool in handling data visualization and analysis . It has received high ratings by publications such as PC Magazine, Windows Magazine, Mac User and American Demographics Magazine.

The most significant difficulty in developing the cost analysis software has been in interfacing it with Maplnfo 3 .0. Most of the coding in the cost analysis software has consisted of interfacing Maplnfo and the software that will do the actual transportation cost calculations .

C. Proposed Programming Language

Three programming languages have been considered for possible use for the software development. These programming languages include Borland C++, Visual Basic 3 .0, and MapBasic.

1. Borland C++

In the last several years, the C programming language has become the overwhelming choice of programmers to implement sytems. C is unique among programming languages because it provides the convenience of a higher-level language such as BASIC or Pascal, while allowing the close control of a computer's hardware and peripherals that assembly language does . It is also believed that C is more portable than most other languages, that is, a C program can be easily converted to run on different computers .

There are two important factors that make Borland C++ undesirable in developing the cost analysis software. The first is that the proposed software will have to be easily interfaced with the Maplnfo 3.0, which is a windows application software. Difficulty is anticipated in completing such an interface by using Borland C++ .

A second factor is even greater difficulty in interfacing Borland C++ with a Windows

application. The interface requires using a programming technique called Application Programming Interface (API). This would require using shared Dynamic Link Libraries (DLL) files to obtain windows libraries to interface with Windows applications . Borland C++ does not have C code to link the DLLs like Visual Basic does. A DLL can only be created by the Borland C++ compiler.

A DLL is a file that contains a set of objects and specialized functions that Windows programs can call on as needed. It allows any Windows applications to call the libraries of external routines (Maplnfo Data Set files). Windows application supports procedures collectively called Dynamic Data Exchange (DDE) which will be required to automatically transfer data from one windows application to another. Borland C++ is not specially designed to make Windows interfaces with the user. Therefore, using Borland C++ will increase the complexity of the required software interface .

2. Visual Basic

Visual BASIC is a popular programming application that has come into use in the last five to six years. It is especially made for Windows applications . It includes graphical design tools

and a simplified high-level language . All procedures and functions utilized in the Windows

environment are available to the user through special designated commands. These allow the developer to take full advantage of the windows operating environment, graphical environment,

and the operating systems. The dynamic data exchange (DDE) commands provided with Visual Basic to interface with Maplnfo is much simpler in comparison to the API calls required by Borland C++ . The visual interface with the user by using Visual BASIC programming language is considerably simpler than Borland C++, because all the window objects are available in Visual BASIC. These include dialog boxes, message boxes, icons, label, check box, option button, scroll bar, command buttons and etc. Visual BASIC is a programming language that is centered on two types of objects, windows and screens, which can be easily created by the developer. The applications that utilize Visual BASIC are graphically efficient and can include all of the most common features expected by the users in a Graphical User Interface (GUI) environment . Visual BASIC also offers a highly flexible response to mouse and keyboard events at run time, a powerful math and string-handling library, easy to use string variables, both fixed arrays and dynamic arrays, and random access and sequential file support. Maplnfo is a windows application and will heavily interface with the cost analysis software. Thus, the entire package can be written asa Windows application.

3. MapBasic

Maplnfo has its own programming language called MapBasic [32] . MapBasic is designed to work with Maplnfo. It is a structured BASIC-like language that enables both experienced and novice programmers alike to create powerful new desktop mapping applications . Maplnfo

concepts are imbedded within this language . MapBasic contains a set of Structured Query Language (SQL) functions that can be used to access Maplnfo data or Data Sets . Other useful applications that MapBasic offers to the user are as follows :

• Create custom Maplnfo applications.

• Perform sophisticated geographic SQL queries .

• Create custom menus and dialog boxes with Title Option, Buttons, Check Boxes, Pop-up menus, etc.

• Automate repetitive operations, and much more.

MapBasic is a complete programming language built around the Maplnfo Graphical User Interface (GUI) products (Windows, Macintosh, Sun, HP, and Intergraph). It significantly enhances productivity when creating custom desktop mapping applications . Also, compiled MapBasic applications are completely portable across all GUI platforms, while protecting the original source code. With MapBasic, applications can be created on one platform and run on all other GUI platforms without changes to the original program. SQL commands can be incorporated into MapBasic programs to automate powerful new geographic extensions such as Within, Intersects, and Contains. MapBasic contains complex relational database queries that can be performed with just one command . For example, queries can be made to select all states that intersect with a specific highway . MapBasic has a geographic analysis tool that enables the user to add graphical objects to maps, create buffers, find perimeters and distances .

In the Microsoft Windows environment, both MapBasic and Maplnfo support Dynamic Data Exchange (DDE) conversations, which allow other Windows programs to communicate with Maplnfo and also enable MapBasic applications to communicate directly with other Windows applications. For example, an Excel macro can send a ZIP Code taken from a spreadsheet cell to Maplnfo to determine which territory the ZIP Code resides in . The territory

can be read back to Excel. When the user chooses two points on the map, the DDE reads and

sends the chosen point in distance form to the program to calculate the cost . _{MapBasic supports} DDL's in Windows, which allows Maplnfo for Windows applications to call libraries of external routines. Therefore, MapBasic was initially considered as an alternative choice as a programming language in the event Visual BASIC proved to be unacceptable .

D. System Requirements

This section presents the requirements for executing the software being developed . The software is designed to be executed on a personal computer. Visual Basic is a programming system for Windows applications . It has been used in developing the Transportation Cost Analysis software. The standard edition of Visual Basic requires an IBM-compatible computer with an 80486 platform. Other system hardware and software requirements include:

• A hard disk with at least 150 megabytes

• A mouse or other pointing device compatible with mouse or supported by Microsoft Windows

• A graphics quality super VGA monitor

• At least 8 MB ofRAM(Random Access Memory)

• A CD ROM disk device

• Microsoft (MS) -or PC-DOS, version 3.1 or higher

• Microsoft Windows 3 .1 or higher, running in standard or enhanced mode

• Transportation Cost Software Architecture

The software is divided into three major components . These components are Maplnfo, the main calculation program, and the data files. Figure 1 shows the block diagram representation of the general structure of the software and the information flow between components . The first component is Maplnfo which acts as the server. For Maplnfo, the user needs to open the workspace called INTMOD.WOR which will be used to visualize the selected region . The second component is the main program that calculates transportation costs _{. The data is sent} from Maplnfo to the main program by a DDE conversation .

MAPINFO SERVER

II DISTANCE DDE CONVERSATION

MODEL COST ANALYSIS CLIENT SOFTWARE

DATA FILE DATA FILE 12

Figure 1: General Flow Chart of Software Data Flow

A DDE works in the following manner. One Windows application (called the client) tells another Windows application (called the server) that it wants information . For example, suppose the distance in miles is needed in determining the true cost in cost analysis software . The software will have to request it from Maplnfo . This can only be done by using a DDE link . Technically, a DDE conversation takes place between two windows applications on the same

machine with both applications supporting these conversations. In a DDE conversation, one application is active and will begin the conversation (client) . The client application takes all the initiative. f\For example, it sends instructions and queries to the server application. In the proposed conversation, Maplnfo is the server and the transportion analysis program is the client .

Before the conversation takes place, the main program opens a conversation to initiate the conversation between main program and MapInfo . This conversation stays active throughout the program. When the program needs any information about the distance traveled from MapInfo during the program, the main program writes a code which initiates and requests MapInfo to send the data to the main program . A command of "terminate" or "terminate all" is initiated by the main program when closing one or all open DDE conversations. The DDE conversation always automatically updates the information in the developed program after the initial exchange of data. Therefore, the first task for the developed program initially set up the memory data exchange which will be ready to use when needed.

The third component consists of two data files . The first data file consists of the cost per ton-miles data that is obtained from published rates . The first is a comparative cost assessment of different transportation modes using published transportation cost rates . The second data file will contain the data for the societal cost calculation, i .e. the true cost per ton-mile . The second is a similar comparative assessment which considers societal costs . These costs consider federal and state contributions to the transportation infrastructure that are not reflected in published transportation cost rates. The data stored in these two files is sent to the main program when needed. The sending procedures are carried by initiating a "request" command.

F. Organization of the Transportation Cost Analysis Program

The program does not require the user to input a large number of parameters . The first parameter that is needed from the user is the selection of the line segment of the transportation path by using SQL-Select. The SQL-Select is a function that can be used to access Maplnfo data or data sets. In this case, SQL-Select is referred to the selection of transportation routes or line segment of the selected path. The second parameter that the user needs to be specified is the weight of the commodity being transported and the chosen transportation mode .

The program is user friendly. For example, step by step instructions and help menus are provided throughout the program, if further information is necessary . The software is stored on a hard disk which has an executable file. The user needs to load Maplnfo 3 .0 and open a file in Maplnfo in order to run the program . As soon as the user loads these files, a map of the United States will be shown on the screen.

1. Principal Software Modules

The developed transportation cost analysis program is divided into three main modules embedded within four to five functional levels of the system . All modules are interrelated to one another. Data entries for each level need to be correctly entered in order for the software to proceed to the next level. The program has an introductory module which provides the user with an overview of the software. Figure 2 presents a general flow chart of this introductory module.

2. Initial Operation

The first level of the system will present an introduction to the software and calls the modules within the second level of the system. The developed program's icon is shown on the screen at the beginning of the software . If the user clicks on the program's icon, a welcome

START

MORE
INFORMATION

NO

CONTINUE

YES

COMMODITY
WEIGHT SELECT
PATH

CALCULATION

YES

Figure 2: Information Flow of Introductory Module

screen will appear, and three small boxes which represent three alternatives will be shown on the

screen. These alternatives consist of the following :

(1)

Continue -Lets the user proceed to the next level of the application, which is the input area.

(2)

More Info -This selection gives the user some additional information and also a detailed description of the project.

(3)

Quit -Allows the user to exit the program.

Figures 3 and 4 present the different levels of the software and the introductory information flow chart of the developed program .

LEVEL 1: INTRODUCTION TO THE COST ANALYSIS PROGRAM

LEVEL 2: INPUT AREA

TRANSPORTATION COST ANALYSIS SOFTWARE

LEVEL 3 : COST ANALYSIS CALCULATION

LEVEL 4 : SUMMARIZATION OF THE OUTCOME & RESPECIFICATION

Figure 3: Four functional levels of the Program

r --,-\

START

QUIT

MORE
INFORMATION &
DESCRIPTON OF THE
SOFTWARE

Figure 4: Information Flow for Software Introduction

3. Region Selection

The second level of the system has a module assigned as an input area. At this stage the

user will be asked to select his/her analysis region of choice .The whole U. S. map will be

displayed by Maplnfo, and the user will need to select his/her region . However, the only region

that will be valid in this stage is Region 1 which includes Arkansas, Alabama, Tennessee,