Guidebook for Intercity Passenger Rail Service and Development (2016)

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

B-1 As described in several sections of the main guidebook text, the Passenger Rail Investment and Improvement Act of 2008 (PRIIA) included multiple mandates to improve planning, operation, and safety of intercity passenger rail services in the United States. PRIIA Section 209 specifically and subsequent implementation has mandated that sponsoring states or other sponsoring enti- ties must reimburse Amtrak (or an alternate provider) for the full cost of their respective short- distance services. PRIIA Section 209 regulations also specifically allow states and other sponsors to unbundle various aspects of passenger rail service and contract with independent entities for specific services. To help states reach agreement with Amtrak on costs to be charged if purchased from Amtrak, PRIIA Section 209 also mandated that an improved cost allocation model be developed and applied to help determine more accurate costs for sponsor reimbursement. The complexity of this costing formula, related issues, and the resultant implications for each of the states or sponsoring entities has led to substantial concern and potential dispute by several of the states in the initial years of application. This appendix presents some of the fundamental areas of dispute and creative solutions achieved by some of the sponsoring entities in their contracts with Amtrak. Primary Findings Grandfathered 403(b) States Are Among the Most Sensitive Prior to the initial actual implementation of PRIIA Section 209 on October 1, 2013, several key states providing Amtrak-operated, state-sponsored service enjoyed a “grandfathered” status for their historically 403(b) services. As background, the original Rail Passenger Service Act (enabling Amtrak, the National Railroad Passenger Corporation) included a Section 403(b), which specified terms under which states or other government entities could contract with Amtrak and pay part of the avoidable loss in a cost-sharing arrangement. This percentage of avoidable loss defined the subsidy requirement for a particular state’s service. Over the years until PRIIA 209, Amtrak had been allowed to determine specific state-by-state agreements on the percentage of coverage, with several states at least initially covering as little as 50%. Separately, PRIIA 209 mandated a consistent and uniform application of its new full-cost formula, so several states did not have to pay any ongoing operating subsidies for some existing intrastate service because these trains were historically considered elements of the basic national Amtrak system. However, several of these unique circumstances were offset by a quid-pro-quo in which these states made substantial capital investments in fixed facilities and/or rolling stock for these routes, comparable to or even greater than the equivalent 403(b) avoidable loss operating subsidy requirement. A p p e n d i x B PRIIA 209 Cost Formula Transparency, Costing Granularity, and Related Issues

B-2 Guidebook for intercity passenger Rail Service and development Examples of trains that states had previously enjoyed without any direct operating subsidy requirements are as follows: • Many of New York State’s New York City–Albany–Niagara Falls Empire service frequencies. • The three Michigan Chicago–Detroit–Pontiac Wolverine service trains. • Three of Illinois’ Chicago–Springfield–St. Louis Lincoln service trains. Because the new PRIIA Section 209 formula now clearly (and equitably) puts all short-distance trains (under 750 miles) into the financial responsibility of the sponsoring states, application of the formula places a greater incremental burden on those states that had enjoyed a “free ride” than the many new-start state corridors that always required a percentage copayment for operating losses. Two examples of services where the states/supporting agencies have a history of pay- ing only partial costs are the Northern New England Passenger Rail Authority’s (NNEPRA’s) Boston–Portland–Brunswick Downeaster serving Massachusetts, New Hampshire, and Maine; and California’s Capitol Corridor Joint Powers Authority (CCJPA) Sacramento–Oakland– San Francisco Capitol Corridor trains. From multiple interviews and inferences, states facing the largest increase in annual subsidy requirements in the recent PRIIA implementation restructuring have been among the most sensitive, and frequently, most vocally concerned. Instituting a Costing Formula Common to All State/Operating Entity Customers Is a Fundamentally Fair Principle, but Unintentionally Resulted in Winners and Losers Most states would objectively agree that initiating a uniform, predetermined, unbiased costing formula for use by all states and other operating entities (e.g., joint powers authorities or multi- state entities) is a fair and equitable principle. Unfortunately, in practice, this concept has led to some states feeling they are “losers” in relative terms when compared to other states, which they perceive as “winners.” For those states that had long-standing established Section 403(b)/ state-sponsored service relationships with Amtrak, it was common practice to negotiate specific deals unique to that state/entity (some involving proprietary agreements) annually or biannually. States that operated a combination of historically national system and historically 403(b) services had similar annual (or occasionally multiyear) costing agreements. The new PRIIA formulaic cost assignment process announced in 2010 (and developed jointly by Amtrak, a working group of affected states, and the FRA) for full initial implementation on October 1, 2013, allows limited or no individual flexibility—there are no more special deals. Several of the states involved in the working group as well as the larger States for Passenger Rail Coalition were frustrated with the process leading to the first-year implementation and thought the initial agreement did not necessarily meet their individual best needs. Use of an Allocation Model Based on Prior Cost to Forecast Future Cost When Anticipating a Significantly Changed Level of Service Raises Concerns Both in the initial interview process and follow-up targeted synthesis discussions, a cost estimating topic that arose multiple times in various contexts was the use of the new Amtrak Performance Tracking (APT) Cost Allocation model to estimate future costs for service, especially in those cases where the state or purchasing entity was planning to adjust the service level in the upcoming year. Amtrak and the purchasing states recognize that, even though APT represents a clear improvement in allocation method over its predecessor Route Profitability System (RPS) and resultant improvement in cost accuracy, it is still a (prior) cost allocation model rather than

pRiiA 209 Cost Formula Transparency, Costing Granularity, and Related issues B-3 a (future) cost forecasting model. Several states believe their future financial needs planning could be better met by a specific model focused on forward-looking forecasting rather than back-looking allocation. This is of particular concern among states planning for frequency or larger service-level changes under consideration for the upcoming year. Cost Categories with Broad Acceptance Experience to date indicates a limited number of cost categories, especially among “direct costs,” that receive broad general acceptance. Because key direct costs (e.g., fuel and power, train and engine [T&E] crew, onboard service [OBS] labor, and host railroad access fees) can at least be relatively simple and clearly historically assigned to routes and specific trains, these categories generate the least frustration or disagreement between Amtrak and state customers in cost contract negotiations. The general sense of agreement is limited primarily to (relatively) stable, unchanging levels of service, where the forecast (anticipation) of future year’s cost is based on the prior year’s cost plus and unit-cost-based or known labor-agreement-escalation-based increase. As with the other cost categories where there is greater general debate, the states are not comfortable with Amtrak’s ability to accurately forecast direct costs for substantial changes in levels of service (e.g., planned increase or decrease in frequency). Cost Categories Generating Conflict These contentious cost categories are mostly those based on unit-use allocation and cost out- puts of Amtrak’s relatively new APT cost allocation model. Some of the specific issues raised by interviewees about several of these cost categories are described below. Shared-Station Costs There are two fundamental concerns the interviewed states and purchasing entities expressed about their allocated amounts of shared-station costs. The first concern, applicable to stations served only by Amtrak trains, is the output variable (e.g., passenger trips) used to allocate and split costs between the subject state-sponsored trains in the state contract from other trains at that station, especially if some of them are basic-system long-distance trains that involve more costly use-per-person of station facilities (e.g., if checked baggage is used). The second concern, applicable primarily to large urban stations that serve non-Amtrak commuters, is the predetermined split of the overall total station cost between Amtrak-passenger-related and commuter-passenger- related use. Amtrak believes (and states do not generally contest) that the recently introduced APT cost allocation model does a reasonably good job of distributing actual station costs to specific Amtrak trains using the station, including algorithms to differentiate use by short-distance and long- distance passengers. Conversely, the larger and contentious issue of cost distribution occurs at large, joint-use stations, among the most complex of which is New York’s Penn Station (heavily used by Amtrak, New Jersey Transit [NJ Transit], and Long Island Railroad [LIRR] passengers). The gross total Amtrak cost for the station is predetermined through contractual agreements between Amtrak and the several commuter systems sharing use of the station. It is that (predetermined) total Amtrak dollar amount that is then used as input to APT, which then allocates the state- portion vs. the national-system portion. Affected states in this situation perceive that there may be issues in the larger (predetermined) split between commuter agencies and Amtrak rather than the APT split between long-distance intercity and state-sponsored, that could result in the states having to pay more than they believe to be an accurate and fair share.

B-4 Guidebook for intercity passenger Rail Service and development Recognizing the complexity of the distribution formulas, and the observation that there is no clearly right or wrong answer, it is likely that the most satisfying result the states are likely to achieve will be through continued detailed interactive discussions and negotiation with Amtrak as the shared-station cost figure is developed for each annual (or longer term) PRIIA 209 operating contract. Commissary/OBS Supplies Costs With the few exceptions discussed elsewhere, most state-sponsored trains use Amtrak- provided food and beverage service and are charged for the commissary and OBS supply costs. For this category, there appears to be little argument among the parties over the relative allocation of cost to a particular train or service. The concern widely expressed in state interviews was the wish that Amtrak management could either find a way to reduce the overall total commissary/ OBS supply cost, or conversely, improve the food/beverage product quality to help justify its high cost. Marketing/Advertising/Sales Promotion Costs Some states and purchasing entities, especially those that have been providing state-sponsored services for many years, have gradually developed their own successful and independent advertis- ing and promotional programs for their trains. These can include locally printed and circulated expanded timetables, promotional flyers, and co-advertising with other entities. The perception at some of these states is that they should not have to additionally contribute (financially) to Amtrak’s general marketing and awareness campaigns. Other states, including some of the newer and smaller scale operators, do not choose to do their own promotion and have no generic problem with the cost associated with marketing and sales promotion. In these states’ case, the biggest concern is that Amtrak may not provide enough focus on their particular service to have a meaningful positive effect on ridership. Reservation and Sales Office Call Center Costs Akin to the observations on state concerns about shared-station costs, the primary issue of concern with the allocation of reservation and sales office (RSO) costs to states is that of proper split between (presumed simpler and shorter) calls related to short-distance, state-sponsored trains and (longer, more complicated) calls for overland/long-distance trains. Although some state-sponsored trains offer (slightly enhanced) business class cars, most of the services are a simple, one-class Point A to Point B ticket, with relatively simple, quick script of availability and fare, possibly ending with making a reservation. Conversely, RSO calls for long-distance trains can be complex, involving multiple choices of routing, types of onboard space (e.g., describing the attributes of different types of sleeping car rooms), day of week (for less-than-daily services), and so forth. Despite Amtrak’s RSO call center time-use surveys, and resultant adjustments to the APT allocation, some of the states still think they are overcharged for this function. Police Costs Although this is a relatively small actual cost item, the topic is raised because of the varying com- ments received from various state sponsors during the interviews. Several operators of relatively lightly patronized rural trains along routes with minimal station facilities and little expectation of crime commented that they rarely (if ever) saw an Amtrak police officer and did not perceive any need for this service, and therefore did not think they should be charged. Conversely, other operators believed that they need and value any degree of security/protection they can procure and gladly pay for the provision of Amtrak police services.

pRiiA 209 Cost Formula Transparency, Costing Granularity, and Related issues B-5 General and Administrative/Overhead Costs Treatment of overhead and general and administrative (G&A) costs in practically every allocation formula (historic or forecasting) is frequently an area of dispute and debate. In the context of being inherently contentious, the treatment and resultant cost in the Amtrak PRIIA 209 formula does not induce an undue amount of concern for most states. Notwithstand- ing, several states or contracting organizations indicated a strong desire for greater transparency and an annual recalculation of the respective percentages, in the expectation that more elements of Amtrak’s overall G&A/overhead costs would be entirely “off books” to state-sponsored, short- distance trains. Equipment Provision and Maintenance Costs For the many states that have chosen Amtrak to provide rolling stock and motive power for their services (as opposed to purchasing or leasing their own), one of the most significant ongoing cost categories is the Provision of Equipment (Capital Recovery) and Maintenance category. In essence, this account becomes a quasi “rent,” given that Amtrak owns the pool of rolling stock, and states effectively are required to compensate Amtrak for the prior purchase (capital depreciation) and the ongoing maintenance of equipment (MOE) cost. Proper costing of the depreciation and/or opportunity cost of state use of Amtrak rolling stock is a complex topic that would require separate economic analysis. A few states have noted that the combined equipment cost accounts for roughly half of their total PRIIA 209 payment to Amtrak. The new APT cost allocation model in use attempts to be much more understandable to state customers and deploys direct assigned unit-use to allocate equipment cost to trains and routes. Many states are relatively comfortable with its transparency but not necessarily with the actual dollar outcome. Other states have difficulty with both the logic and the ultimate resultant cost. Amtrak is correct to point out that it provides a positive extra service by having a larger com- mon equipment pool available, supported by this Provision of Equipment (Capital Recovery) and Maintenance charge. If an unexpected last-minute failure should be experienced by a small state-owned dedicated fleet, alternative equipment might not be available to be pressed into service, but (at least at the larger Amtrak facilities) Amtrak would be able to provide an emer- gency substitution. Similarly, Amtrak’s pool would likely be better positioned to meet extra requirements for weekend or holiday peaks. A Few States/Operating Entities Have Resolved Problems with a Particular Cost by Either No Longer Providing a Service Element or Procuring from Alternate Sources As of the end of the first full year of PRIIA Section 209 “full-state-funding” service contracts, no state had substituted outright an alternate qualified provider for Amtrak for its basic operation. One state was close to confirming a change of basic provider as of the beginning of the second (fiscal) year, October 2014, but various issues postponed this implementation until sometime in 2015. Conversely, at least a half-dozen states/contracting entities have chosen to seek an alternate provider for one or more specific cost categories, taking advantage of the unbundled procurement approach, or they have chosen to not offer (and hence not purchase) service for a select function. Alternate Provision of Food and Beverage Service In large part because the NNEPRA’s Downeaster is a stand-alone service route (not directly connected to other Amtrak service) operated out of Boston’s North Station, and a newly added

B-6 Guidebook for intercity passenger Rail Service and development state-sponsored train, when introduced in the early 2000s, it was able to avoid Amtrak commis- sary and OBS labor costs by directly purchasing service from an independent local provider. Consistent customer survey responses and general comments show that this has worked well for the NNEPRA, providing a somewhat more locally focused range of products, and at retail prices slightly lower than for comparable items in Amtrak-operated cafe cars. NNEPRA Management note that, although this is a negative concession (i.e., the NNEPRA must pay an incentive to the independent food and beverage (F&B) provider for this service), it still costs the agency less than a comparable service provided by Amtrak and arguably results in greater customer satisfaction. Such a service, although hypothetically possible under the PRIIA Section 209 rules, would be difficult (if not impossible) to re-create on a route that had historically provided food and beverage service by Amtrak using Amtrak Service Workers Council (ASWC) unionized OBS employees. One attempt to reintroduce food service formerly provided by ASWC on Empire Service trains between New York and Albany almost a decade ago failed in less than 1 week. Amtrak attempted to deploy staff and supplies from a nationally well known, outside independent sandwich chain as vendor to provide the service, but ASWC opposition was so vocal and effective that the demonstra- tion was abandoned. Discontinuing Provision of F&B Service For a few select very-short-distance routes, some states have chosen to not offer any onboard F&B service to reduce total subsidy requirement and avoid the commissary and OBS labor costs. Examples are the WisDOT–sponsored Milwaukee–Chicago Hiawatha service trains; the INDOT-sponsored Indianapolis–Chicago Hoosier State service (until the recently implemented INDOT-sponsored contract for private, independent provision of premium OBS); and New York State’s New York–Albany Empire service trains. The shortest of the Michigan-sponsored trains operating Chicago to Grand Rapids had done the same for several years, but restored basic AmCafe service to improve customer satisfaction and provide consistency across state-sponsored Amtrak service. Providing (Limited) F&B Service Using Vending Machines Although wholly unsuccessfully when tried elsewhere at least twice in the past, the state of North Carolina appears to have found a way to provide reliable (if somewhat limited) F&B service on its Charlotte–Raleigh Piedmont service trains. Because of the limited menu options provided and the lack of direct labor, this service may be the lowest cost option of any non-zero F&B service. Providing Independently Sourced Call Center Service The CCJPA in California has uniquely (among contracting entities) chosen to provide its own call center service, using dedicated staff of the jointly administrated agency, Bay Area Rapid Transit (BART). This selection, in conjunction with the decision to keep all Capitol Corridor trains unreserved, has significantly reduced the CCJPA’s cost by eliminating Amtrak- provided information and reservation services requiring Amtrak RSOs. Amtrak has observed that there is still some potential for confusion (CCJPA information has a different phone num- ber than the standard Amtrak 800 number) and a small number of calls must be redirected (one way or the other) (e.g., if a CCJPA Capitol Corridor customer wants to connect to another Amtrak route). States Purchasing Their Own Equipment By procuring their own (purchased or leased) equipment, a few states have avoided the some- what contentious quasi-rent Amtrak charges for use of its pooled equipment for state-sponsored

pRiiA 209 Cost Formula Transparency, Costing Granularity, and Related issues B-7 service. Among the most substantial and long-standing independent equipment arrangements is the State of Washington and State of Oregon joint ownership of dedicated Talgo Trains for service in the Eugene–Portland–Seattle–Vancouver Cascades service. The states have a long-term agreement with Talgo, including a full, high-standard, preventive maintenance program, allowing virtually all of the fleet to be available for revenue assignment at peak times. Although the first generation of these highly popular, European-styled, tilting trains required FRA waivers because of their lighter weight and buff strength, the most recent version is fully FRA compliant. Another example of state-owned rolling stock, in this case also including motive power, is the NC DOT’s dedicated fleet for the Charlotte–Raleigh Carolinian service. NCDOT purchased and then fully remanufactured to its own specifications a small, dedicated fleet of secondhand passenger coaches and locomotives. In doing so, it has controlled costs and also have been able to reflect their unique identity in a special paint scheme and customized interior configuration (including cafe cars with the above-discussed unique onboard F&B vending machines). There is also a recently implemented INDOT-sponsored contract for private, provision of rolling stock for the Hoosier State service. The Recent Deterioration of On-Time Performance (OTP) of State-Sponsored Services Is a Growing Concern States and sponsoring agencies have always been concerned about their relatively limited ability to influence and help improve the OTP of the services they purchase from Amtrak. Some have worked jointly with Amtrak and host operating freight railroads to identify and then fund worthwhile spot improvements to relieve bottlenecks and chokepoints. Some larger state programs have developed ongoing joint OTP analysis and oversight teams with Amtrak and respective host freight railroads to review performance and identify ways to address problems quickly, either through schedule adjustments or spot capital improvements. Although states/contracting agencies had already been concerned by and involved in OTP management, there has been a significant recent increase in the need for proactive steps as unprecedented freight traffic growth has led to substantial deterioration in passenger train perfor- mance and timely arrivals. Furthermore, some of the states have noted that they now feel an even greater need to take action because they have a larger direct financial stake in the operation, given the full operating loss requirement imposed by PRIIA Section 209. The unrelated coincidence of significantly increased freight traffic/decreased passenger OTP and states newly assuming 100% of their services’ operating losses has put extra pressure on rail managers to demonstrate proactive involvement to help solve the problems and find ways to improve their OTP. Conclusions The evolution of the annual/biannual PRIIA Section 209 contract negotiations between states and Amtrak (or potentially other providing operators) is a work in progress. The process for the second full year, which took effect October 1, 2014, although still contentious, was achieved somewhat more smoothly than the first year and the more recently concluded agreement taking effect October 1, 2015, resulted in even less contention. Amtrak is attempting to provide more transparent and detailed information to its state customers, while the states are becoming better prepared to ask more specific questions as they understand the process better. Among the clearest and most universal messages from states and purchasing entities is the desire to understand Amtrak’s APT more fully and to work with Amtrak to modify some specific allocation formulas to address specific concerns. States believe they would benefit from a new variable operating cost (and revenue) forecasting model for what-ifs and requested hypothetical

B-8 Guidebook for intercity passenger Rail Service and development changes in level of service on specific state-sponsored routes. Notwithstanding, Amtrak can and does perform this analysis for specific (large) changes when requested by states, (e.g., for outright new routes or non-negligible increases/decreases in train frequency). It is also possible that FRA, in its dual role of Amtrak oversight and co-negotiator in Amtrak/ state contract partnerships, could independently commission development of new cost forecasting models, in much the same way as it contracted with the Volpe Center to develop the APT cost allocation model to replace the RPS for route accounting.