Review and assessment of the New Jersey freight assistance program

Research in Transportation Economics xxx (xxxx) xxxx

Contents lists available at ScienceDirect

Research in Transportation Economics journal homepage: www.elsevier.com/locate/retrec

Review and assessment of New Jersey rail freight assistance program Kaan Ozbaya, Bekir Bartinb,∗, W. Bruce Allenc, Shrisan Iyerd, Martin Robinse, Marc Weinere, Hani Nassiff a

C2SMART Center, Department of Civil and Urban Engineering, New York University, 15 MetroTech Center, 6th Floor, Brooklyn, NY, 11201, USA Department of Civil Engineering, Altinbas University, Mahmutbey Dilmenler Caddesi No: 26, Bagcilar, Istanbul, Turkey c Business Economics and Public Policy, Regional Science, and Transportation, University of Pennsylvania, 1407 Steinberg-Dietrich Hall, 3620 Locust Walk, Philadelphia, PA, 19104, USA d C2SMART Center, New York University, Six MetroTech Center, Room 401, Brooklyn, NY, 11201, USA e Rutgers University Bloustein School of Planning and Public Policy, 33 Livingston Ave, New Brunswick, NJ, 08901, USA f Rutgers University, Civil and Environmental Engineering, 623 Bowser Road, Piscataway, NJ, 08854, USA b

A R T I C LE I N FO

A B S T R A C T

Keywords: Short line railroads Rail freight assistance Application scoring IMPLAN

Class III railroads, also called “short line railroads” are crucial to economic activity and transport of goods. The private and social benefits of maintaining short line railroads is well-established in the literature, yet the cost of doing so is often prohibitive for short line operators. Thus, several assistance programs have been established, most of which are operated by the state departments of transportation. This review of these programs indicated that there is no unique methodology for administering rail freight assistance programs and evaluating applications submitted for funding, and that most of them lack clear administrative details such as published project criteria, ranking procedure and post evaluation requirements. The objectives of this paper are to present a comprehensive review of the New Jersey rail grants program, identify the shortcomings of its evaluation process and provide recommendations for improving the program's project selection criteria and application ranking. It is found that the current scores used in the evaluation process cannot always clearly differentiate between applications for funding. To that end, using the results of a targeted survey of experts, an adjustment to the current review criteria and the corresponding scores is suggested for implementation.

JEL classification: L92 R42 O18

1. Introduction In the United States, railroads are designated in three classes by the Surface Transportation Board (STB) based on revenue, as follows (STB, 2017): ⁃ Class I: Carriers with annual carrier operating revenues of $467 million or more ⁃ Class II: Carriers with annual carrier operating revenues between $37.4 and $467 million ⁃ Class III: Carriers with annual carrier operating revenue of $37.4 million or less, and all switching and terminal companies regardless of operating revenues. Class I and Class II carriers are referred to as national and regional railroads, respectively. The Staggers Act of 1980 deregulated the American railroad industry and led to national railroad companies

selling or abandoning rail tracks that were less profitable and had low density per mile. Much of this track was taken over by Class III carriers, also known as “short lines” (Miller & Stich, 2014). Short lines connect freight origins or destinations to Class I long-haul rail freight lines. According to the Federal Railroad Administration (FRA) report, the number of short line railroads has more than doubled since the Staggers Rail Act, from about 220 companies in 1980 to more than 540 in 2014 (FRA, 2014). Although they operate at a relatively small scale compared to the nation's major long-haul freight railroads, short lines are crucial to economic activity and transport of goods as they are the linchpin of the railroad network, efficiently connecting many communities to the nation's mainline railroad system. Despite their economic significance and value to the freight transportation system, most short line rail companies struggle to be financially viable because their current revenues are not sufficient to overcome the backlog of deferred maintenance on their lines that had occurred under the mainline railroads' ownership (Sage,

∗

Corresponding author. E-mail addresses: [email protected] (K. Ozbay), [email protected] (B. Bartin), [email protected] (W.B. Allen), [email protected] (S. Iyer), [email protected] (M. Robins), [email protected] (M. Weiner), [email protected] (H. Nassif). https://doi.org/10.1016/j.retrec.2019.100742 Received 24 October 2018; Received in revised form 23 July 2019; Accepted 29 August 2019 0739-8859/ © 2019 Published by Elsevier Ltd.

Please cite this article as: Kaan Ozbay, et al., Research in Transportation Economics, https://doi.org/10.1016/j.retrec.2019.100742

Research in Transportation Economics xxx (xxxx) xxxx

K. Ozbay, et al.

more likely to be able to operate low-density lines profitably, ⁃ Short lines can provide superior shipper service, especially for local demand, compared with Class I railroads or trucking companies that might operate in the same local markets, and ⁃ If short lines are successful, they reduce the number of truck shipments, ultimately reducing highway congestion, maintenance and rehabilitation costs.

Casavant, & Eustice, 2015). The sustainability of short line railroads depends on two areas of concern (Llorens & Richardson, 2014). The first concern is their ability to shift from 263-kip (263,000 pounds) loaded car weight to 286-kip loaded car weights, following the lead of Class I railroads. Short line railroads interchange with Class I railroads, and to remain viable short lines need to upgrade to handle heavier cars (Casavant & Tolliver, 2001). However, many short lines do not have the underlying track and structures capable of supporting these cars (LaDOTD, 2003). Carrying 286-kip cars on short lines with light rail and poor tie and/or ballast conditions poses a safety problem due to their increased vertical and lateral loading of the track, which may result in derailments even at low operating speeds, as explained in detail in Zarembski (2014). The second concern is their ability to receive funding for major rehabilitation projects. It is evident that these two concerns are interrelated and pose a unique dilemma for these rail carriers. Therefore, in order to preserve the railroad system's network and operations, many states have programs that provide loans or grants for rehabilitation and improvement projects. State support programs for short lines can be categorized into four different forms (Qiao et al., 2016): (1) grant programs (2) loan programs, (3) loan/grant hybrid programs and (4) tax-based incentives and benefits. Federal support is also provided through Railroad Rehabilitation and Improvement Financing (RRIF) program, 45G short line tax credit, and occasionally the Transportation Investment Generating Economic Recovery (TIGER) grant program (Ozbay et al., 2015, Miller & Stich, 2014). There are also some loan programs that allow for conversion to a grant if performance targets, such as jobs or local carloads, are achieved (Sperry & Morgan, 2013). In New Jersey, the New Jersey Rail Freight Assistance Program, henceforth referred to as “rail grants program”, awards approximately $10 million annually in grants to freight rail operators for acquisition assistance, construction, rehabilitation and demonstration type projects. The purpose of this program is to promote and sustain economic development, as well as to maintain a balanced transportation system, where rail is used in lieu of trucks when economically viable. The review process for the selection of candidate projects for funding is made based on the information on each project, their cost estimates and the estimated value added to the state. The financial viability of any state funded projects should be evaluated by identifying, estimating and comparing their costs and benefits. The ultimate goal is to allocate society's resources efficiently and ensure that any proposed project promises returns to society greater in value than its costs. To that end, the main objective of this paper is to review the NJ's rail grants program as a case study, identify the shortcomings of its evaluation process and provide recommendations for improving its project selection criteria and application ranking. The following tasks were conducted towards this objective: (1) the relevant literature and the available information on state funding programs were reviewed, (2) the historical project application data obtained from the NJDOT were analyzed, (3) interviews were conducted with industry experts, transportation agencies, railroad carriers, (4) a targeted survey of experts including rail grants evaluators and applicants was conducted and (5) an adjustment to the current review criteria and corresponding scores to improve the evaluation process was suggested. In addition, an analysis of a selected set of projects submitted to the program was conducted, in which the estimates of their economic benefits obtained from IMPLAN, an input-output economic modeling tool, were compared with those in the project application forms.

The benefits of short line railroads can be categorized as economic and transportation related. The economic benefits connote to not only the direct jobs that short line railroads create, but also the indirect jobs and positive induced effect leading to more employment opportunities and economic activities (Ozbay et al., 2015; Qiao et al., 2016). Several studies found that negative consequences of railway abandonment include increased transportation costs and slowed economic activity and growth due to the fact that freight rail provides shippers with cost-effective transportation, especially for heavy and bulky commodities (Allen, 1975; Babcock et al., 1992, 1995). Freight transportation can be a critical factor in retaining and attracting industries that are central to state and regional economies. Qiao et al. (2016) estimated the economic impact of 35 short line railroads in Texas and reported that their operations contribute to nearly 1500 jobs and $355 million dollars in economic output at the state level. Feser and Cassidy (1996) performed an ex-post evaluation of rail investments and highlighted that there are many benefits to preserving local railways in addition to increased economic development, such as job creation, shipper impacts, highway damage, and safety. They concluded that the transportation related benefits in terms of reduced maintenance and operation costs were the most overlooked benefits of railways, and that the evaluations of infrastructure investment focused too heavily on job creation. Short line railroads benefit the transportation system because they provide shippers with an alternative to trucks. This reduces the number of truck trips, which in turn reduces the pavement deterioration caused by heavy trucks. For example, Babcock and Sanderson (2006) found that short line abandonment in Kansas would lead to $58 million in road damages because of increased trucking. Warner and Terra (2005) estimated that if the short lines were not upgraded and the goods formerly transported on rail were moved to trucks, the pavement damage over a seven-year period would total $49.8 million for rural interstates or $226.1 million if they used only rural major collectors. Activity on short-line railroads was also shown to have positive effects on Class I railroads, increasing their cost savings by 23% (Tolliver, 1989). Transportation related benefits of short lines also include lower congestion and truck-related crashes, environmental health and safety benefits such as fuel efficiency and less air pollution, and safer shipment of hazardous materials (Qiao et al., 2016). The analyses conducted by Qiao et al. (2016) indicated that short line railroads have significant advantages over trucks and that 14 short lines in Texas contributed to 12 million dollars in savings in 2015 in terms of reduced congestion, accidents, emissions, operating and maintenance costs. On the other hand, there are several arguments to be made against short line railroad funding (Babcock et al., 1993): ⁃ Short lines are not likely to survive in the long run because of large, deferred maintenance expenses, ⁃ Short lines are too dependent on a few commodities for most of their revenue, and ⁃ Short lines are too dependent on Class I railroads for equipment and market access.

2. Literature review

Indeed, there are benefit-cost studies that have shown the cost of maintenance due to the abandonment of some railways far outpaces the costs imposed by equivalent trucking trips. For example, a North Dakota study by Bitzan and Tolliver (2001) estimated impacts to highways from the conversion to trucking trips to be $1 million—but the cost to upgrade the rail infrastructure would be $191 million.

According to Babcock, Russell, Prater, and Morrill (1993), the advantages of short line railroads are as follows: ⁃ Short lines have lower labor costs than Class I railroads, and are 2

Research in Transportation Economics xxx (xxxx) xxxx

K. Ozbay, et al.

two major freight rail programs (Cambridge Systematics, 2011): the Rail Freight Assistance Program (RFAP) ($10.5 million annual program) and Capital Budget Transportation Assistance Program (TAP) ($20 million annual program). The Pennsylvania Rail Freight Assistance Program uses the Pennsylvania Rail Benefits Estimator (PRBE), a Microsoft Excel spreadsheet-based model, to evaluate grant applications and conduct cost-benefit analysis. The considered benefits include reduced highway maintenance costs, lower highway safety costs, and lower emissions resulting from reduced truck traffic. The tool also estimates the direct and indirect effects on employment, including shortterm (construction) and long-term (railroad) jobs created or maintained. The key input is car loadings for truck-rail diversion. Qualitative scoring criteria include several categories: infrastructure, coordination, economic, environmental sustainability, safety and security, financial and institutional, benefits of truck reduction, types of benefits, and track condition. Each category contains several qualitative measures assigned a numeric score by the PennDOT review staff. The qualitative scoring sheet sums the scores to provide a composite qualitative score. Finally, using the inbuilt factors and multipliers, the PREB uses the quantitative input and qualitative assessment to score the project. Post-evaluation of projects is conducted, where applicants are expected to achieve 50% of the projected carloads after 5 years. The Florida Department of Transportation (FDOT) selects freight rail projects for state funding from its surface transportation program and Strategic Intermodal System (SIS) program. Short line rehabilitation projects are required to have a 25% match from the applicant. The FDOT evaluates the project proposals using SIS goals and uses the Florida Rail Investment Calculator (FRIC) to carry out a cost-benefit analysis (FLDOT, 2006). When evaluating applications, FRIC considers the following benefits: (1) transportation impacts: reduction in highway maintenance and reduction in costs for shippers, (2) economic impacts: job creation, tax revenue, and reduction in passenger delays, and (3) external impacts: land use, safety, security, and environmental impacts. Once the benefits are quantified, they are combined, considered, and discounted over a total project span to compute cost-benefit ratios. The Iowa Rail Revolving Loan and Grant Program (RRLG), administered by the Office of Rail Transportation through the Iowa DOT, receives appropriations for loans and grants. RRLG focuses on two distinct evaluation processes for different kinds of projects: (1) targeted job creation and (2) rail network improvement, such as service improvements, rail yard expansion and improvement, rehabilitation, and industrial park development (Iowa DOT, 2011). To encourage companies to locate or expand in Ohio, the Ohio Rail Development Commission (ORDC) offers loans and grants and may issue bonds for qualified rail projects (ODOT, 2010). Grants are also limited to projects with significant job creation or retention (25 + jobs), and clawback is employed when creation/retention numbers are not met. The loan program provides a five-year loan with a two third prime interest rate. Benefit analysis is often used to determine eligibility for assistance. Eligible benefits include job creation and retention, transportation cost savings, new investment, increased viability of rail, relief of highway congestion and maintenance, safety

Among national studies, the National Cooperative Highway Research Program (NCHRP) Report 586 – Rail Freight Solutions to Roadway Congestion presents guidance on evaluating the potential feasibility, cost and benefits of investing in rail freight solutions to alleviate highway congestion from heavy truck traffic (NCHRP, 2007). It summarizes some key observations drawn from national case studies on the approaches of public agencies to rail freight benefits analysis: ⁃ There is no single methodology for assessing projects with multiple categories of benefits and costs that include weighting factors varying from agency to agency. ⁃ Equity and political issues involved in infrastructure financing must be considered. ⁃ All approaches to benefit estimation are subject to debate concerning which types of projects are being analyzed and how publicprivate projects should be structured. ⁃ Cost-benefit analyses must show that total project benefits exceed total costs, using the time value of money to compare the current and future costs and benefits. Cost-benefit analysis should also include non-monetary aspects, which must be quantified and weighted according to agency needs and considerations. Based on the majority of the literature and the industry experts who were interviewed as a part of this paper, it can be claimed that there are private and social benefits of maintaining short line railroads, but the cost of doing so is often prohibitive for short line operators. Thus, several assistance programs have been established—most of which are operated by state DOTs. 3. Review of Other State Programs Many state DOTs have rail grant programs designed to help fund local rail improvement projects to maintain or grow rail freight. Several states such as New York, Minnesota, and Illinois, utilize loans to further extend limited funding, and negotiate loan terms and matching requirements based on the ability of the applicant or funding situation. Clawback provisions are used by many programs, few of which are in conjunction with mandatory post-evaluation. Post-evaluation requires data collection for several years following construction completion to ensure that projected carloads, job creation/preservation, and the critical components of cost-benefit analysis are not unrealistically inflated. The projections must be realized, usually to a certain minimum level, or some or all of the grant money must be returned. While a cost-benefit analysis is required by all states, the details for each analysis vary, with some states, such as Oregon and Virginia, outsourcing their economic analysis to third parties. A summary of the characteristics of rail grant programs of the reviewed states is given in Table 1. The following subsections briefly explain the cost-benefit methodologies and monetization factors used by select state rail freight assistance programs. More detailed information on these programs can be found in Sperry and Morgan (2013), Ozbay et al. (2015) and Qiao et al. (2016). The Pennsylvania Department of Transportation (PennDOT) funds Table 1 Comparison of state support programs. State

NJ

Grant Only Grant & Loan Match required Clawback/Collateral B/C and/or Economic Analysis Independent Economic Analysis Post-Evaluation

X X X

NY

PA

MN

X X X X X

X X X

FL

ID

IL

IA

KS

X X X X

X

X X

X

X X X X

X

X

MD

X X X X X

X X

ME

OH

X

X

X

X X

X

Note: Maryland's program provides loans/grants to private operators of state-owned lines, with terms on a case-by-case basis. 3

X X X X X

OR

VA

X

X

WA

X

X X

X X X X X

X

Research in Transportation Economics xxx (xxxx) xxxx

K. Ozbay, et al.

administrative details such as published project criteria or ranking procedure. For example, it was found that 42 percent of the programs did not have any published information on the evaluation criteria. Among programs with published criteria, approximately half of them included a published formal scoring method. The next section presents an overview of the NJ rail grants program, pinpoints some probable flaws during the project selection process, and recommends ways to improve them.

improvements and more. The Ohio DOT uses the following percentagebased scoring criteria to evaluate grant applications: Transportation Factors (55%), Community Economic Growth and Development Factors (25%), and Local and Private Investment Factors (20%). To assess these impacts, inputs are added to a spreadsheet-based tool to conduct costbenefit analysis. Several monetization factors are utilized and built into the tool. The Oregon Department of Transportation (ODOT) administers ConnectOregon, a lottery-based bond program to improve highway, air, and rail infrastructure (McMullen, B. S., 2010). The program was authorized in 2005 and has since funded $300 million in projects, nearly 50 percent of which are rail projects. Applications are reviewed by stakeholders, transportation experts, and local residents, and then approved by the Oregon Transportation Commission. In lieu of projections from the railroads/applicants, an independent consultant calculates the economic benefits of each application. There are two freight rail funding programs administered by the Washington State Department of Transportation (WSDOT): the Washington Rail Bank ($5 million every two years) and the Freight Rail Assistance Program ($2.75 million per year) (WSDOT, 2017). Because the state constitution prohibits transfer of public funds to private enterprises, WSDOT cannot provide funds directly to railroads. Therefore, the applicants for freight rail funding assistance are municipalities, public agencies, ports and other special districts. WSDOT uses a spreadsheet-based cost-benefit analysis (CBA) methodology to evaluate applications. The State Department of Commerce provides input on approximately 25% of the evaluation having to do with job creation and other economic variables. Depending on the nature of the application, other state agencies are involved (WSDOT, 2008). A review of these programs indicated that they have several common elements, but they also differ in some characteristics and criteria for disbursing funding, revealing the following key results:

4. Review of NJ rail grants program The evaluation process of the NJ rail grants applications is based on information provided by the rail carrier, such as the project description, cost estimates, number of customers, number of carloads served, revenue per carload, projections for future car loadings, number of jobs estimated to be saved or created, and a description of how the project benefits the State's rail system and its transportation goals. This information is used to rate the grant application in the following categories defined by the NJDOT: (1) Economic benefits to the state, (2) efficient and responsive freight distribution, (3) energy and environmental factors, (4) highway congestion mitigation, (5) benefits to applicants and (6) benefits to the community. In addition, some of this information is also used by the NJDOT in a cost-benefit spreadsheet. The benefit-cost ratio of applications is one of the secondary criteria, used only if there is a tie between applications in the primary (essential) criteria scores, as explained in the next section. The data gathered for this paper include (1) electronic records of the rail grants applications received by the NJDOT, and (2) scorecard evaluation scores given by reviewers. The electronic records provide basic information on the applications including the project name, project number, the requested funding, whether the project was funded or not, the state share of cost, municipalities, agreement dates and invoice information. Analysis of the application records indicated that between 2003 and 2011 the rail grant program approved 143 of 242 candidate projects. In addition, the NJDOT paid an average of 78% of the project cost, which totaled $68 million of the $88 million costs incurred by the 136 funded programs. Since 2003, the NJ rail grants program has received an average of 26 applications per year, and on average, has funded 16.5 applications annually. The funding levels since 2003 have fluctuated between $7 million and $14.5 million per year, with an average funding of $10.5 million. It should be noted that the number of funded projects has steadily declined while the funding level has remained stable. This suggests that the program is now funding fewer projects with a higher level of support.

⁃ There is no single unique methodology used to administer rail freight assistance programs and evaluate applications submitted for funding. ⁃ Several states solely offer loans or use loans in conjunction with grants to make use of limited funds. Some percent of project cost is often matched by applicants. Matching percentages differ between project types and states. Several states also negotiate or offer higher match contributions based on the ability of applicants and project needs. ⁃ States tend to measure similar types of public benefits, mostly focused on quantitative measures such as rail traffic (carloads) and job creation (short- and long-term). Network improvements, connectivity, and other, less precise measures are also considered, as are safety and pollution impacts. ⁃ The factors used to weigh project evaluation criteria differ from state to state but in all cases, a positive cost-benefit ratio is required for funding eligibility. ⁃ Most states require applicants to provide basic information regarding a forecast of the expected shift in freight rail traffic, and job preservation/creation. Some states post-evaluate projections to ensure they are not inflated, and clawback provisions are usually employed to recover loan/grant money from the applicants if projections are not met. ⁃ Some states employ outside consultants to conduct the benefit-cost analyses, while others utilize other state agencies involved in economic development. ⁃ Funding agency staff often makes site visits to project locations and applicants make in-person presentations to agency staff. Such practices help application evaluators better understand the purpose and need for the proposed projects.

4.1. Application scoring Before 2009, grant applications were reviewed by a panel of NJDOT staff and prioritized, and an award list was decided based on the application information and public input. The program implemented a more structured and formal evaluation process in 2009, in which evaluation scorecards were used to rate each application on the project's ability to improve the economy, freight efficiency, and environment. Evaluation scorecards were filled out by the NJDOT reviewers based on essential and secondary criteria. The current essential and secondary criteria used in the NJDOT application scorecard are summarized in Table 2. Note that the scorecards were completely overhauled after the first year (2009) and modified in 2010 and 2011, but many of the original criteria remained intact. It should also be noted that the secondary criteria scores are only used in the event of a tie in the essential criteria scores. Our analysis of the scorecards indicates that in 2009, an application could receive at most 39 (maximum 13 essential criteria points per reviewer times 3 reviewers), and the average winning application received 33.6, which is 4.9 points higher than the average for all

In addition, as revealed by the review conducted by Sperry and Morgan (2013) a high proportion of these funding programs lack clear 4

Research in Transportation Economics xxx (xxxx) xxxx

K. Ozbay, et al.

Table 2 Current NJDOT application scorecard. Criteria Essential Criteria Economic

Freight Distribution Energy and Environment Highway Congestion Mitigation Secondary Criteria System/safety Continuation of a Previous Project Benefit Cost Ratio

Quality Assurance Applicant Performance

Definition of Categories

Points

S1

S2

S3

Does the project maintain existing jobs? Does the project support new jobs? Does it increase the number of businesses served by rail? Will the project improve customer service? Will the project increase service reliability? Does the project reduce truck emissions and fuel consumption? Does it improve the safety of transportation of hazardous materials? Does the project reduce long haul trucking?

0–2 0–2 0–3 0–2 0–3 0 or 1 0 or 1 0 or 1

1 0 2 1 1 1 1 1

1 1 1 1 2 1 1 1

2 1 1 2 3 1 1 1

Is the project a systematic or line improvement, which improves the state's rail system? Is this a continuation of a partially funded rail grants program project from a prior year? How strong is the benefit cost ratio? Ratio between 1 and 5.9: 1 point Between 6 and 9.9: 2 points; Above 10: 3 points Applicant provides a clearly defined scope of work and cost estimation Has this applicant met previous requirements for program implementation? Does the applicant have a history of executing agreements within 6 months and preparing error-free invoices?

0–3 0 or 2 1,2, or 3

1 1 2

0 1 2

2 0 2

0–2 0 or 1

1 1

1 1

1 1

Economic criteria can be tested on the basis of the historical data of jobs created and maintained by the customers served by the respective railroad. In the Analysis section of this paper, a comparison of some of the economic criteria predicted by the applicant and the ones estimated by IMPLAN, an input-output economic modeling tool, is presented. It is shown that the predictions by applicants are not always consistent with the model's estimates. Among the project selection criteria listed in Table 2, benefit-cost ratio stands out as the one that requires quantitative input. The NJDOT utilizes a standard spreadsheet for the benefit-cost analysis for each application which takes into account project costs, predicted number of new permanent jobs, revenue per carload, current and new car loadings, discount rate, miles of track and manufacturing salary. These numbers are self-reported by the applicants and the validity of the estimates is not investigated before using them in the calculation of a benefit-cost ratio. Some other possible improvements can be made to the scoring method used by the NJDOT. For example, double counting is an issue, with some factors considered both in the essential criteria and then again in the benefit-cost calculation. Job related questions in the Economic section under essential criteria are based on maintaining or increasing numbers and can only be scored as 0, 1, or 2; no weights are given to percentage growth or total numbers. Yet these values are used in the benefit-cost spreadsheet, the results of which are used only when there is a tie in the essential criteria total scores. Additionally, it may be redundant to ask if the project reduces trucking emissions as well as if the project reduces long-haul trucking. The following questions arise after evaluating the weights and factors considered by the rail grants program and reviewing the current practice by other state DOTs:

applications (28.7). In 2010, the highest possible score for the essential criteria was 51 (maximum 17 essential criteria points per reviewer times three reviewers), and the average winning score was 33.8, 2.2 points higher than the average score of all applications (31.6). In 2011, the highest essential criteria score possible was 45 points (maximum 15 points per reviewer times 3 reviewers), and 33 points for secondary criteria (11 points per reviewer times 3 reviewers on average). The average winning essential criteria score in 2011 was 26.4, 3.7 points higher than the average score of 22.7. 4.2. Scoring related issues The scorecards play a particularly important role in the evaluation process. Table 2 shows the scores given to each category by three different reviewers for one of the projects in 2011. It is evident that the scores for each criteria category vary significantly from one reviewer to another. The scores given to some of the essential criteria questions, such as “Will the project increase service reliability?” differs greatly between reviewers. Similarly, the scores given to the question “Is the project a systematic or line improvement, which improves the state's rail system?” in the secondary criteria, varies considerably between reviewers. The substantial discrepancy between the reviewers’ scores is observed in other applications as well. To evince this issue better, let us focus on the average of the highest deviation from the mean score for the essential criteria for each application in 2011. For example, in Table 2, the total scores for each reviewer are 14, 14 and 18; with a mean score of 15.33, and the highest deviation from the mean is 2.67 points. This value reaches to as much as 5.10 points across three reviewers with an average of 2.59 for all applications in 2011. As discussed in the next subsection, even though using scorecards is a more systematic and formal means of evaluating rail grants applications, using a more quantitative analysis to determine evaluation scores can potentially lessen these variations in scoring.

⁃ Should economic justification be given greater weight? o If a project shows promise for inducing productive investment and job creation/retention in NJ, then the amplitude of points available should be expanded. ⁃ Should applications that seek improvements enabling a short line to carry 286-kip carloads be given high priority? o Rail cars with a gross weight of 286-kips are becoming standard in the U.S. This improvement could affect the ability of a short line to retain its customers and thus substantiate its future economic viability. Adding this improvement could then become a priority category. ⁃ Should the amount of financial assistance requested always be fully granted? o In some cases, other sources of financial assistance can be

4.3. The shortcomings of the evaluation process Each criterion should be scored with a standard method and the documents supporting the scoring of each criterion by individual reviewers should be carefully reviewed for any errors. Also, instead of scoring projects individually by each reviewer, a collective decision can be made by all following a thorough discussion on each application and analysis of the benefits associated with it. This practice might avoid scoring discrepancies. For example, the criteria based on energy and environment requires some kind of practical air emissions analysis. 5

Research in Transportation Economics xxx (xxxx) xxxx

K. Ozbay, et al.

Table 3 Survey results. Essential Criteria

Old max points

Eigenvalues from AHP

Scaled

New max points

The project maintains existing jobs The project supports new jobs The project increases the number of businesses served by rail The project improves customer service/increases service reliability The project reduces long haul trucking, emissions, and fuel consumption The project improves the safety of transportation of hazardous materials Secondary Criteria The project is a systematic or line improvement which improves the state's rail system The project continues/builds upon previously funded improvements The project has a high benefit/cost ratio The project upgrades load rating to 286 kips

2 2 3 5 2 1

0.12 0.12 0.11 0.09 0.09 0.09

3.08 3.11 2.85 2.21 2.34 2.39

3 3 3 2 2 2

3 2 3 –

0.09 0.11 0.10 0.08

2.28 2.86 2.52 1.92

2 3 3 2

Since its inception, AHP has been used as a decision support and evaluation methodology in a multitude of applications including transportation policy and planning (Bagchi, 1989; Banai, 2006; Khasnabis & Chaudhry, 1994; Levine et al., 1999; Tabucanon & Lee, 1995; Zak, Fierek, & Kruszynski, 2014). In order to translate the judgments from a paired comparison, the AHP fundamental scale for pairwise comparisons suggested by Saaty (1980) is used. Available values for the pairwise comparisons are members of the set: {9, 8, 7, 6, 5, 4, 3, 2, 1, 1/2, 1/3, 1/4, 1/5, 1/6, 1/7, 1/8, 1/9}. Intensity of importance is measured by 1, 3, 5, 7 and 9 for equal, moderate, strong, very strong and extreme importance, respectively. Intensities of 2, 4, 6, and 8 are used to express intermediate values. The pairwise comparisons are then arranged in a matrix. For example, if there are n number of criteria that are compared pairwise, an n by n matrix is arranged with its upper (or lower) diagonal elements include the comparison scores, and diagonal elements equal to 1. If criterion x is viewed as being equal in importance as criterion y, then x is assigned the numerical value of 1. If x is viewed as being extremely more important than y, it is assigned a numerical value of 9. Likewise, if x is viewed as extremely less important than y, then x is assigned a numerical value of 1/9. This numerical scale allows the decision maker to quantify the intensity of the judgments while introducing a mathematical basis for the overall analysis. The survey questions are taken from the NJDOT application scorecards. The questions are limited to ten, with some combination and elimination of criteria in agreement with the NJDOT personnel, as shown in Table 3. 10 responses are received: 5 NJDOT respondents, 2 other DOT respondents, 2 anonymous and 1 railroad member. Once all of the surveys are completed, the numerical values for each question are averaged among all survey participants and then the weights are calculated using the AHP method. The differences in scoring point values versus the current NJDOT scoring value weights are also shown in Table 3. It is interesting to note that following the AHP process, the old and new maximum scoring points are quite similar for most of the questions in the survey. The customer service and service reliability criteria, for example, is reduced from a total of 5 points to 2 points. It is a subjective criteria, thus reducing its weight highlights the importance of choosing more quantitative ratings. However, given the small sample size, these new points are not recommended for adoption at this point. Instead, it is recommended that a similar exercise with a larger group of participants be conducted by the NJDOT to improve the current evaluation process. The main problem in doing this is the difficulty in finding knowledgeable panel members who can provide input relevant to NJ specific questions. Another recommendation based on the outcome of this effort is the need for the consolidation of the scoring criteria to the 10 questions listed in the survey to avoid double counting by asking questions dealing with a single issue more than once. The AHP approach presented above can aid the department in

suggested and/or encouraged. ⁃ Should tax revenues be included in benefit/cost calculation? ⁃ Should monetization of emissions, safety, maintenance, and congestion costs be included in both the benefit/cost calculation and other project evaluation criteria? 4.4. How to improve the evaluation process? To address the above listed questions, a total of 16 DOT/agency program managers from several states are contacted, including Ohio Rail Development Commission, Florida DOT, Virginia Department of Rail and Public Transportation, Oregon, Maryland, Minnesota, Idaho, as well as six NJ railroad managers and railroad industry experts and two consultants. Some of the information obtained from these telephone interviews is presented in the Review of Other State Programs section of this paper. In-depth telephone interviews are conducted with several DOT/agency program managers, consultants and industry experts from New York and Pennsylvania, as these two states border NJ and have a profound reciprocal impact on the NJ rail system. 4.4.1. Interviews with consultants/industry experts Select short line railroads in NJ who responded to inquiries were visited for personal interviews. Several railroads provided data on their operations and traffic spanning the past several years. They also detailed their experiences with the NJDOT rail grants program. The interviews provided key carload traffic data from railroads as well as several key suggestions for improving the program. Many suggested that the program could benefit from more rigorous financial analysis of the applicant railroads and further studies should investigate how to accomplish this. During the interviews, close communication was established with the NJDOT program managers to share ideas and discuss the implementation of potential program changes. As a result of these exchanges, the NJDOT expressed interest in a more rigorous evaluation of applicants, including finances and an in-person application presentation. In addition, while the NJDOT does not post-evaluate grants at the present, it was expressed as a possibility for the future. The department is open to anything allowed by the statute and administrative code governing the program. Some practical problems are noted with regard to implementing clawback, such as the lack of manpower/expertise. 4.4.2. Surveys A survey of rail grants evaluators and applicants is conducted which focuses on the criteria used in the project evaluation process. The survey employs the Analytic Hierarchy Process (AHP) created by Thomas L. Saaty in the early 1970's, which was inspired by mathematics and psychology (Saaty, 1980). AHP provides users with a rational framework for making a complex decision and evaluating numerous alternate solutions. It gives the user the ability to derive ratio scales from pairwise comparisons. This allows the participants who are decision makers/experts to incorporate judgment into the decision. 6

Research in Transportation Economics xxx (xxxx) xxxx

K. Ozbay, et al.

than a combination of projects 4 and 5, awarding projects 2 and 4 would have potentially yielded more jobs according to IMPLAN. This is an important observation that demonstrates that the use of a sophisticated economic forecasting tool such as IMPLAN can yield valuable information. The estimated numbers of jobs created as a result of the rail grants applications are also investigated. Specifically, the number of jobs estimated by IMPLAN and the ones predicted in the applications of nine funded projects in 2010 are compared, as shown in Fig. 1. As seen in the figure, the number of jobs predicted by the applications is much larger than the numbers estimated in IMPLAN. When comparing the two, the focus is on the total number of jobs created irrespective of job type, because jobs estimated by IMPLAN are mostly indirect jobs and not direct jobs, whereas the predicted jobs in the applications are mostly in the manufacturing category, i.e. direct jobs. The jobs created by activities in IMPLAN are based on the multipliers with respect to the dollars invested. Indirect jobs are important elements to consider while deciding which projects will be funded. Therefore, information from IMPLAN or other transportation input-output models can be considered alongside information relating to the indirect jobs created along with the direct jobs estimated by the applications, and other project related information. However, the results obtained from IMPLAN should be used cautiously. Although it is a popular input-output tool, it has certain limitations. For example, IMPLAN multipliers reflect industry linkages in a local economy at a given time, and it does not account for changes in consumer or industry behavior based on a direct effect. In addition, time required for total economic impacts to be realized is not specified (IMPLAN, 2017). Finally, as mentioned in the Literature Review section, short line railroads can result in transportation-related benefits in terms of reductions in congestion, vehicle crashes, emissions and pavement damage. A comprehensive evaluation process should include the estimation of these benefits as well. These estimations would assist the reviewers in calculating a more accurate benefit-cost ratio as well as in answering qualitative questions such as “Does the project reduce truck emissions and fuel consumption?“, “Does it improve the safety of transportation of hazardous materials?” or “Does the project reduce long haul trucking?” as shown in Table 2. Although the cost of abandoning short line railroads can be estimated, as reported in the literature, it is not always straightforward to estimate the project-based transportation related benefits. The primary reason is the fact that most projects are small in scale. For example, in NJ the average funding allocated per project is approximately $500K without the match, which is rather small compared to other transportation investments. In addition, the scopes of most projects are rehabilitation and acquisition assistance, which do not have direct and easily quantifiable transportation impacts.

selecting the most appropriate benefit criteria and determining their corresponding scores. Further improvements to the current evaluation process should include the assessment of the economic impact of each project application. As stated earlier, states tend to measure project benefits similarly, mostly focusing on benefit-cost ratio and employment creation. Similarly, the NJDOT's current evaluation process relies on the estimates of the new and sustained employment figures, provided by the applicants. However, as with most states, there is no mechanism in place for controlling the validity of these estimates. To that end, the next section presents an alternative method for estimating the economic benefits and selecting candidate projects. 5. Impact of economic benefits on application ranking Like NJ, most states only consider benefit-cost ratio to be one component in the overall analysis. While some state DOTs, including the NJDOT, utilize in-house spreadsheet tools to estimate this ratio, others, such as Oregon and Virginia, outsource their economic analysis to third parties. Currently, the NJDOT does not require the use of an economic analysis tool, such as IMPLAN, RIMS II or REMI model in the evaluation process. For this paper, IMPLAN (IMpact Analysis for PLANning), an input-output economic modeling system, is selected to test its effectiveness on quantifying the economic impacts of projects supported by NJ's rail grants program. IMPLAN is a widely used, nationally recognized economic impact model, first developed by the U.S. Forest Service (AKRF, 2013). Many state and local governments use IMPLAN to estimate the economic changes caused by local economic activity. An example of such activity can be an extension of a railroad connecting the industrial park to the main rail line. This new construction could transfer more goods via rail, which is expected to boost local businesses. IMPLAN can be used to provide an estimate of the new level of overall economic activity. The indirect economic activities associated with the original economic activities can also be estimated by IMPLAN. If a business purchases goods at another local industry, the business is supporting the local industry and IMPLAN estimates all levels of economic activity supported by that business. The impact of a project, in this case a transport by rail industry project, is obtained in the form of employment, income, value added, and total output values for each industry. Tax impacts and project impacts over various future years can also be determined in IMPLAN. In order to observe the cumulative effect of a series of projects, multiple projects can be assigned to a particular county. The most efficient combination of the projects in a particular county can be obtained using a trial and error method, observing the changing employment numbers in different combinations of projects. For example, five projects that were in consideration for rail grants in Morris County, NJ in 2010 are selected for this analysis. Their cumulative effect in IMPLAN is analyzed using the detailed project information obtained from the electronic records and scorecards. The results are shown in Table 4. Note that the estimates of their direct impacts are not included because they came out to zero for all five projects. The average scores given by the reviewers are also updated using the estimated values from IMPLAN and shown in the table. According to the historical data, only projects 4 and 5 were funded in 2010. Their cumulative effect can be found by analyzing the scenario with a combination of these two projects. When the cumulative effect of projects 4 and 5 is calculated, it can be seen that total employment, labor income, value added and output are 5.5, $376.733, $533,934 and $843,515, respectively. To determine if the combination of the selected projects is the optimum choice, the scenarios with a combination of two or three different projects at one time can be analyzed. Using the values presented in Table 3, it is observed that the combination of projects 2 and 4 yields total employment, labor income, value added and output as 7.7, $530,554, $749,125 and $1,187,926, respectively. Hence, rather

6. Summary and conclusions The majority of the literature and the interviews with industry experts confirm the private and social benefits of maintaining short line railroads, but the cost of doing so is often prohibitive for short line operators. Thus, several assistance programs have been established—most of which are operated by state DOTs for rail lines within their states. There is no unique methodology used to manage rail freight assistance programs and evaluate applications submitted for funding. As discussed in Sperry and Morgan (2013), although these programs provide the much needed funding for short line railroads, most do not have clear selection criteria and a transparent administrative decision making process. Therefore it is necessary to improve the project evaluation and selection process to allocate public resources efficiently, and ensure that the funded projects offer greater values than their costs. In this paper, several state DOT programs are reviewed and the 7

Research in Transportation Economics xxx (xxxx) xxxx

K. Ozbay, et al.

Table 4 IMPLAN combined scenario results for projects in morris county, NJ in 2010. Project No

Cost

Impact Type

Employment

Labor Income

Value Added

Output

Average Score

1

$996,518

$1,138,850

3

$677,500

4

$1,585,500

5

$349,133

2.1 0.8 2.8 2.3 0.9 3.2 1.4 0.5 1.9 3.3 1.2 4.5 0.7 0.3 1

$149,273 $44,830 $194,103 $170,594 $51,233 $221,826 $101,486 $30,478 $131,964 $237,424 $71,304 $308,728 $52,298 $15,706 $68,005

$193,943 $80,124 $274,067 $221,644 $91,568 $313,211 $131,855 $54,474 $186,329 $308,474 $127,440 $435,914 $67,948 $28,072 $96,020

$318,026 $116,576 $434,602 $363,450 $133,226 $496,675 $216,216 $79,256 $295,471 $505,833 $185,418 $691,251 $111,421 $40,843 $152,264

10.88

2

Indirect Induced Total Indirect Induced Total Indirect Induced Total Indirect Induced Total Indirect Induced Total

11.13

8.50

9.88

10.63

Fig. 1. Comparison of jobs estimated by IMPLAN and predicted by nine NJ rail grants applications in 2010.

of the program is conducted based on the historical project application data obtained from the NJDOT. The data gathered for this paper include (1) electronic records of the rail grants applications received by the NJDOT, and (2) scorecard evaluation scores given by reviewers. The review of the available scorecards from the previously funded and non-funded applications reveals that the ratings derived from current evaluation criteria cannot always clearly differentiate between projects. More specifically, the current structure produces evaluation results within a small range. There is also potential double counting between the sub-criteria used for the benefit cost analysis and the essential criteria used in the NJDOT scorecard (Table 2). The weights assigned to each criterion in the scorecard could also be revised so they can be used to differentiate between feasible (realistic) and infeasible applications, and also to reflect the priorities of the NJDOT. In order to address these shortcomings, the evaluation criteria are updated based on interviews with industry experts, transportation agencies, railroad carriers, and the review of the relevant literature. In addition, a survey of 10 experts including rail grants evaluators and applicants is conducted, which focused on the criteria used in the project evaluation process. The survey used the AHP method and the analysis of survey data results in different scores for the evaluation

compiled information is summarized in Table 1. The review showed that many states utilize loans to further extend limited funding and negotiate loan terms and matching requirements based on the ability of the applicant or funding situation. Clawback provisions are used by many programs, only a few of which are in conjunction with mandatory post-evaluation. Several states solely offer loans or use loans in conjunction with grants to make use of limited funds. Some percent of project cost is often matched by applicants. Matching percentages differ between project types and from state to state. States tend to measure similar types of public benefits, mostly focused on quantitative measures such as rail traffic (carloads) and job creation (short- and longterm). Network improvements, connectivity, and other, less precise measures are also considered, as are safety and pollution impacts. The factors used to weigh project evaluation criteria differ from state to state but in all cases, a positive cost-benefit ratio is required for funding eligibility. In this paper, NJ rail grants program is reviewed and summarized in detail as a case study. The objective is to present a comprehensive review of the program, identify the shortcomings of its evaluation process, and provide recommendations for improving its project selection criteria and application ranking. To that end, a comprehensive review 8

Research in Transportation Economics xxx (xxxx) xxxx

K. Ozbay, et al.

criteria. However, given the small sample size, these new scores are not recommended for adoption and a similar exercise with a larger group of participants is recommended. The current review process of NJ's rail grants program is based on the project information provided by the rail carrier. Another shortcoming of the current evaluation process is the absence of an administrative control mechanism that can validate the reported benefits of each project by the applicants. An alternative method for estimating the economic benefits and selecting candidate projects is demonstrated. Based on the literature review, it is determined that IMPLAN is a commonly used tool to conduct the economic analysis of investments similar to the ones found in the applications to the rail grant program. Therefore, IMPLAN is used to demonstrate its usefulness as an analytical tool to quantify economic benefits of projects proposed in previous years under NJ's rail grant program. The number of new jobs predicted by the applications and IMPLAN as a result of investments using the funds obtained from the program come out to be relatively different. Moreover, IMPLAN predicts mainly indirect jobs while the applications to the program predict direct jobs. As a result of these findings, it is suggested that, like many other states, the NJDOT can use a combination of IMPLAN and benefit-cost analysis along with other metrics to score the projects. A more comprehensive analysis of the evaluation program should include a comparison of the predicted and actual outcomes of the funded projects. However, NJ's rail grants program does not mandate a post-evaluation. Therefore, the comparison of applicant-predicted with actual results is not available. In addition, a more comprehensive evaluation process should also include emissions, safety, road maintenance and congestion costs. As presented in the Literature Review section, many studies estimated the transportation related benefits of short lines by demonstrating the cost of their complete abandonment (Babcock & Sanderson, 2006; Llorens & Richardson, 2014; Qiao et al., 2016). However, the estimation of transportation related benefits of single projects is not straightforward, as they are usually rehabilitation and acquisition assistance projects, and therefore not included in this paper. Finally, it can be concluded that if some of the findings and recommendations mentioned above are implemented by NJDOT, the current program can be improved to make NJ's transportation system infrastructure more efficient overall.

Transportation Review, 42(3), 149–166. Bagchi, P. K. (1989). Carrier selection: The analytic hierarchy process. Logistics & Transportation Review, 25(1). Banai, R. (2006). Public transportation decision-making: A case analysis of the memphis light rail corridor and route selection with analytic hierarchy process. Journal of Public Transportation, 9(2). Bitzan, J., & Tolliver, D. D. (2001). North Dakota strategic freight analysis - heavier loading rail cars. MPC report No. 01-127.4, mountain plains consortium. Cambridge Systematics Inc (2011). Analysis of public benefits for Pennsylvania freight rail funding. Casavant, K., & Tolliver, D. (2001). Impacts of heavy axle loads on light density lines in the state of WashingtonReport No. WA-RD 499.1. Pullman, WA: Washington State Transportation Center, Washington State University. Federal Railroad Administration (2014). Summary of Class II and Class II railroad capital needs and funding sources: A report to congresshttps://www.fra.dot.gov/eLib/Details/ L16020, Accessed date: 3 March 2017. Feser, E. J., & Cassidy, G. D. (1996). Rethinking state rail policy. Policy Studies Journal, 24(1), 57–73. Florida Department of Transportation (2006). Freight rail funding program. http://www. dot7.state.pa.us/BPR_PDF_FILES/Documents/Research/Complete%20Projects/ Planning/Analysis%20of%20Public%20Benefits%20for%20PA%20Rail%20Freight %20Funding.pdf, Accessed date: 14 April 2016. IMPLAN (2017). IMPLAN website. http://www.implan.com/, Accessed date: 4 February 2017. Iowa Department of Transportation (2011). Rail revolving loan and grant program guidelines. www.iowadot.gov/iowarail/assistance/documents/RRLGProgramGuidelines. pdf, Accessed date: 19 July 2016. Khasnabis, S., & Chaudhry, B. B. (1994). Prioritizing transit markets using analytic hierarchy process. Journal of Transportation Engineering, 120(1), 74–93. Levine, J., Park, S., Underwood, S. E., & Wallace, R. R. (1999). Stakeholder preferences in advanced public transportation system planning. Journal of Public Transportation, 2(2), 25–45. Llorens, J., & Richardson, J. A. (2014). Economic impact analysis of short line railroadsReport No. FHWA/LA.14/527. Baton Rouge, LA: National Center for Intermodal Transportation for Economic Competitiveness. Louisiana Department of Transportation and Development (2003). Louisiana statewide rail system plan. Prepared by Wilbur Smith Associates. McMullen, B. S. (2010). “Multimodal freight investment criteria”. Oregon department of transportation report. http://www.oregon.gov/ODOT/TD/TP_RES/ ResearchReports/Multimodal_Investment_Criteria.pdf, Accessed date: 21 February 2017. Miller, C. R., & Stich, B. (2014). Realizing the economic development benefits of shortline railroads: The case of Mississippi. Regional Science and Practice, 6(1), 1–11. National Cooperative Highway Research Program (2007). Rail freight solutions to Roadway congestion-final report and guidebookReport 586. Transportation Research Board of the National Academies. Ohio State Department of Transportation (2010). Ohio statewide rail plan. Final Report https://www.dot.state.oh.us/Divisions/Rail/Programs/StatewideRailPlan/ Documents/Ohio%20Statewide%20Rail%20Plan%20-%20Final%20Report %20Complete.pdf, Accessed date: 19 July 2016. Ozbay, K., Allen, W. B., Robins, M., Iyer, S., Nassif, H., Weiner, M., et al. (2015). Impact of the rail grants programNew Jersey Department of Transportation Research Report (Report No: FHWA-NJ-2015-002). Qiao, F., Gao, L., Salfarriage, D., You, B., Li, Q., Song, L., et al. (2016). Transportation and economic impact of Texas short line railroadsTexas Department of Transportation Research Report (Report No: FHWA/TX-0-6887). Saaty, T. L. (1980). The analytic Hierarchy process. New York: McGraw-Hill. Sage, J., Casavant, K., & Eustice, J. B. (2015). Washington state short line rail inventory and needs assessmentWashington State Department of Transportation (Report No: WA-RD 842). Sperry, B. R., & Morgan, C. A. (2013). Review of state-level loan and grant programs for local freight rail infrastructure. Proceedings of the 2013 joint rail conference, knoxville, Tennessee, april 2013. Surface Transportation Board (STB (2017). ) website. Federal Register, 79(111), 33257. Tabucanon, M. T., & Lee, H.-M. (1995). Multiple criteria evaluation of transportation system improvement projects: The case of Korea. Journal of Advanced Transportation, 29(1), 127–143. Tolliver, D. D. (1989). The benefits and costs of local and regional railroads. Upper Great Plains Transportation Institute Publication No. 90. Warner, J. E., & Terra, M. S. (2005). Assessment of Texas short line railroads. Presented at 2006 annual meeting of the transportation Research board, Washington DC. Washington State Department of Transportation (2008). State rail and marine Office rail benefit/impact evaluation methodology. Washington State Department of Transportation (2017). Freight rail assistance program website. https://www.wsdot.wa.gov/Freight/Rail/GrantandLoanPrograms.htm, Accessed date: 5 March 2017. Zak, J., Fierek, S., & Kruszynski, M. (2014). Evaluation of different transportation solutions with the application of macro simulation tools and multiple criteria group decision making/ aiding methodology, Vol. 111, Procedia – Social and Behavioral Sciences340–349. Zarembski, A. (2014). Determination of the impact of heavy axel loads on short lines. Delaware Department of Transportation.

Acknowledgements The work presented in this paper was supported by NJDOT and the grant was administered by Rutgers University (FHWA-NJ-2015-002). The authors thank the NJDOT for providing guidance and suggestions at various stages of this study. The opinions and conclusions presented in this paper are the responsibility of the authors and do not reflect the views of sponsors and other participating agencies. References AKRF, Inc (2013). IMPLAN, RIMS-II and REMI economic impact models. http://www.ilw. com/seminars/JohnNeillCitation.pdf, Accessed date: 4 February 2017. Allen, B. J. (1975). The economic effects of rail abandonment on communities: A case study. Transportation Journal, 15(1). Babcock, M. W., Prater, M., Morrill, J., & Russell, E. R. (1995). Competitiveness of short line railroads. Journal of the Transportation Research Forum, 34(2). Babcock, M. W., Russell, E. R., & Burns, R. E. (1992). Economic development and transportation impacts of railroad branchline abandonment in south central Kansas. Kansas Department of Transportation. Babcock, M. W., Russell, E. R., Prater, M., & Morrill, J. (1993). State short line railroads and the rural economy. Kansas Department of Transportationhttp://www.intrans.iastate. edu/reports/rr_rural_economy.pdf, Accessed date: 5 June 2016. Babcock, M. W., & Sanderson, J. (2006). Should shortline railroads upgrade their systems to handle heavy axle load cars? Transportation Research Part E: Logistics and

9