, ,

Letter From Andrew Selden President, United Rail Passenger Alliance

To Hon. Elaine L. Chao, Secretary United States Department of Transportation


Dear Secretary Chao:

URPA has followed with interest the Administration’s budget proposal and its purported impact on Amtrak, as well as Amtrak’s recent statements concerning its results, investment priorities and spending plans. These stories all reflect long-standing misimpressions of both the financial and operational results of Amtrak’s various train operations, and the outcomes of Amtrak’s investment strategies. The budget proposal threatens to eliminate the largest sector Amtrak operates (by output) and its most commercially successful trains (by load factor and market share).

Amtrak’s investment strategies reflect a serious misapplication of scarce capital resources. For decades—subsequent to the departure of W. Graham Claytor as CEO—Amtrak has produced a negative rate of return on invested capital:  Amtrak’s financial results are worse now than they have ever been despite the investment of tens of billions of dollars of free capital in federal subsidies, or more precisely because of how those grants were deployed.

The financial results of operation of Amtrak’s three segments (NEC, Regional Corridors, and Long Distance) are basically the opposite of what Amtrak portrays. Amtrak uses inappropriate metrics that distort the results.

Inappropriate Metrics

We cannot be sure whether Amtrak’s public utterances are designed to mask or place a “positive spin” on its persistently bad results, or instead reflect Amtrak’s own misunderstanding of the metrics used to report its results. Their reports, built on train and segment “ridership,” chronically deceive congress and the media. “Ridership”—the foundation of Amtrak’s self-evaluation–is not a meaningful measure of anything important that Amtrak does. Amtrak exists to move people between cities by rail. “Ridership” does not measure the performance of that mission. It measures only transactions—tickets sold—not the distances over which customers are carried, or the revenues they generate, or the costs required to move them. A static equipment display generates transactions (“ridership”) but zero transportation output.

The only useful metric for the output of a non-commuter, intercity passenger service is NOT “ridership” but “revenue passenger miles” (RPMs) which factors in the distance a passenger is carried. Passenger miles are the output of the enterprise. They are all that matters to an entity whose purpose is to provide intercity passenger transport. Airlines report, rely upon, and plan based upon passenger miles and load factor, not “ridership.” Freight railroads do not measure output by “units carried” but by ton miles of freight transport produced.

Trains Ranked by Output

Amtrak’s segment output of intercity transport may surprise you. In FY2016, Amtrak long distance trains generated 2.649 billion RPMs, while the NEC managed only 1.978 billion, just barely topping the regional corridor trains elsewhere in the country, with 1.899 billion. The long distance trains therefore by a wide margin are Amtrak’s largest segment in producing intercity passenger transport. In FY’15, the NEC had the fewest RPMs and was actually the smallest operation Amtrak had, measured by output. These results are not aberrations, they have been consistent for decades.

And the passenger mile data for the NEC overstates the NEC’s production, because nearly half of its output doesn’t even meet the standard definition of “intercity transport,” consisting instead of quasi-commuter traffic occurring solely within the commuter territory between Philadelphia, New York and New Haven, traffic that could just as well be carried by regional commuter agencies.

The “ridership” data in the NEC is equally misleading for the same reason:  the average trip length in the NEC is so short. About half of all NEC passengers ride solely in the quasi-commuter territory between Philadelphia, New York and New Haven. Under the standard definition employed by the US Bureau of Transportation Statistics, these distances aren’t even “intercity” transport.

Trains Ranked by Load Factor

Another key measure of performance is load factor.

Load factor is important for two reasons. It is a measure of capital efficiency, and it is a direct indication of places where the enterprise is over-invested (low load factor—capacity exceeds demand) or under-invested (high load factor—demand exceeds capacity). Load factor measures how much of its inventory a transport enterprise can sell. It is calculated by dividing RPMs by the number of available seat-miles deployed. Airlines depend upon high load factors, report it, and track it intensely, route by route, flight by flight. Amtrak does not.

Amtrak’s long distance trains consistently have a higher load factor than do the NEC trains, in some instances as much as 20% higher. Amtrak’s segment load factors have been generally consistent for decades. In the NEC, Amtrak’s load factor is just over 50%, for Acela premium services and conventional services alike. Amtrak cannot sell nearly half the NEC’s inventory of available seat miles that it generates each year. This demonstrates objectively and unequivocally that Amtrak is over-capitalized in the NEC because it has far more capacity there than it is able to sell.

In long distance markets, load factors are notably higher. A long distance train is effectively sold out at a load factor of about 65%, and most of the long distance trains approach that level year-round and meet or exceed it in the six or so peak months of the year. (Because of the large number of intermediate stations on a long distance route, a load factor of about 65% is functionally sold out because a seat vacant at any given point already has been sold to a passenger boarding downline. On the western trains, every seat and every berth typically turns over two or more times every trip.)  These load factors objectively and unequivocally demonstrate that Amtrak is under-capitalized in its few long distance markets because demand consistently exceeds supply. These are the only trains Amtrak operates where demand consistently exceeds supply.

Trains Ranked by Market Share

Another key measure of the returns on Amtrak’s use of federal capital is the market share its services have attained.

In the corridors they serve, the long distance trains have a much higher market share for intercity transport than do the trains in the NEC. Amtrak’s market share for intercity transport in the NEC (after 40 years and more than a hundred billion dollars in federal subsidy, in constant 2017 dollars) is well under 2%, which is trivially insignificant to regional intercity mobility in the Northeast. In the markets where they operate, the long distance trains often have shares that approach 5%, more than twice the NEC’s and achieved with sold-out trains and almost no recent federal investment at all.

Amtrak’s claims concerning its modal split with the airlines for end-point traffic in the NEC are meaningless. They are comparing apples to oranges in terms of the traffic flows involved, and a modal split in any case is meaningless in measuring performance when a third modality—motor vehicles—commands a market share greater than 90%.

Based on load factor and market shares, the long distance trains, therefore, are also objectively the most commercially successful trains that Amtrak operates. The load factor numbers are Amtrak’s, and the share values were calculated from data from the US BTS, FAA, AAA and other similar sources.

Amtrak disparages its long distance trains, its largest and strongest product, with the “romance and nostalgia” meme and the assertion that they are merely political sops to non-Northeastern members of congress. In fact, these trains are heavily used by a wide array of travelers. Their annual load factors show trains that are functionally sold out much of the year. The average trip length on Amtrak’s western long distance trains essentially matches the average trip length in domestic commercial aviation, showing that these trains compete effectively in the few markets where the service is offered. That average trip spans three to five meal periods and often involves an overnight segment, supporting the market for sleeping car (first class) services and dining cars. These trains carry every category of passenger as domestic airlines, the only difference being that business travelers are under-represented (but not absent). Just like an interstate highway, they primarily serve the thousands of intermediate origin/destination city pairs in addition to the small proportion of travelers who ride between or beyond  the endpoints.

Trains Ranked by Financial Results

Financial results of operation are another important metric of performance.

Recent media stories have repeated Amtrak’s “big lie” that its NEC trains are, as Business Insider recently wrote, “…Amtrak’s only profitable line.” That is simply a falsehood. Amtrak’s statements and suggestions to that effect silently omit any assignment to NEC trains and their revenues of all or most of the massive annual costs of the fixed facilities of the NEC. (If an airline were to omit its landing fees and terminal gate rentals from its financial results of operation, it would be considered a fraud; Amtrak does that on a daily basis.) These trains may (or may not) cover all of their direct operating costs, but they emphatically do not cover all of the fixed facility costs that they require in order to operate. Those costs (which Amtrak mischaracterizes as “capital” costs) came to about $1.6 billion in FY’16, most of which was spent in and for the NEC, and included another $473 million in deferred maintenance and purchasing (primarily in the NEC) that Amtrak incurred to cover a substantial shortfall in anticipated, budgeted, NEC FY’16 revenues. The NEC once it is charged with its annual fixed facility costs (even after accounting for NEC revenues) appears to account for the major part of Amtrak’s annual loss as shown on its audited financial statements, and to consume 100% of the uncommitted portion of the annual federal subsidy grant to Amtrak.

NEC Deferred Maintenance

Amtrak claims that the NEC requires as much as $24 billion in new federal subsidies to be restored to a “state of good repair.” That is an odd claim for a service that Amtrak claims is “operationally profitable.”

The tens of billions of dollars that Amtrak claims it needs to fix up the NEC is simply the accumulation of decades of deferrals, similar to last year’s $473 million in deferred maintenance and purchasing, to cover annual losses in the NEC in earlier years, taking account of the infrastructure costs along with the operating costs. No reporter or member of congress that we know of has ever asked Amtrak, “If the NEC is so profitable, why does it have billions of dollars of deferred maintenance?” Or, “When you say that NEC trains are ‘profitable,’ does that include the annual costs of the NEC’s track, bridges, tunnels, power stations, passenger stations, signaling, and heavy maintenance facilities?”

Amtrak Misrepresents Financial Results of Operations

Amtrak’s losses cannot, and do not, arise out of its operation of trains other than its NEC trains.

The state-sponsored regional trains Amtrak operates cannot lose money (for Amtrak) because Amtrak operates all of them under contracts that obligate the state to pay Amtrak the difference between the train’s revenues and Amtrak’s “fully-allocated costs” allocated to that train. Under Amtrak’s logic, once a state has paid Amtrak’s “fully-allocated” costs of a given train, by definition no other costs–cash or non-cash—can remain. So, these trains are always cash-positive to Amtrak. They cannot, and do not, require or consume ANY federal subsidy.

The long distance trains also are cash positive to Amtrak. Amtrak said so itself in writing a few years ago, responding to a US Senator’s inquiry as to the impact on the annual subsidy if all the long distance trains were to be eliminated. Amtrak’s answer was that elimination of the long distance trains would increase, not decrease, its annual subsidy need and that its subsequent subsidy need would grow year over year. That can be true if and only if the long distance trains also are cash-positive to Amtrak. Thus, these trains also cannot, and do not, consume any federal subsidy dollars.

That same conclusion can be reached by deduction from the fact that the portion of the $1.6 billion in cash spent predominantly on NEC fixed facility costs and not recovered from NEC train and real estate revenue consumes and accounts for 100% of the unallocated portion of the annual federal subsidy grant to Amtrak. No dollars remain after paying for NEC fixed facility upkeep (not covered by ticket and other revenue) to be available for use on any other trains.

Amtrak’s APT Reports are not Statements of Profit and Loss

Amtrak’s purported statements of train, route and segment performance generated from its “APT” system do not contradict these conclusions. APT does not generate, and is incapable of generating, statements of profit and loss as used in GAAP-compliant financial statements, and its reports do not represent the financial results of operations of any train, route or segment. APT reports are not GAAP-compliant and are not audited.  Rather, they reflect a largely hypothetical, internal, full cost allocation system based on Amtrak management-derived algorithms, and depict only one out of a nearly infinite number of possible cost recovery scenarios to bring Amtrak as a whole to break-even. Amtrak does not have statements of profit and loss compiled in accordance with GAAP for any train, route or segment.

The inescapable conclusion is that neither the regional corridors nor the long distance trains have “losses” that are cross-subsidized by NEC “profits.” That is a myth and a lie. The NEC may (or may not) have an operating margin based solely on its direct operating expenses, but that margin–if it exists—at most goes only part way to covering the NEC’s enormous annual fixed facility costs, but not all of them. THAT is where the annual federal cash subsidy goes.

Failed Investment Strategies

Amtrak’s investment priorities are backwards, and preclude growth and deficit reduction. They produce negative rates of return on invested capital provided free of charge by taxpayers.

Amtrak invests the vast majority of its financial resources into the NEC. Yet, the load factor in the NEC is just over 50%, meaning that Amtrak cannot sell nearly half of its inventory in that market (and never has). That tells us conclusively that Amtrak already is seriously over-invested in that market. It has too much capital deployed in the NEC, because it can’t sell nearly half of the inventory it produces there. (After 45 years and a hundred billion dollars in current dollars, if those seat-miles could be sold, they would be. But they are not.) In the long distance markets, where the trains’ annual load factors are at nearly sell-out levels, Amtrak cannot grow because it has already sold as much of its small and shrinking inventory as is physically possible to sell (and in the sleeping cars, at breath-taking prices). Amtrak is plainly under-invested in these markets because demand exceeds supply for much of the year. Amtrak turns away tens of millions of dollars a year in proffered business that it cannot accommodate for want of long distance carrying capacity. Amtrak’s newest passenger cars on its western trains are more than 30 years old. That’s how long it has been since Amtrak’s last serious investment in these trains or in the scale of their carrying capacity. Amtrak has no plan to expand capacity in these sold-out trains.

The load factors tell a compelling story of misallocated capital, at the same time that they illustrate the absence of any possibility of real growth in the NEC and Amtrak’s willful refusal to invest for growth in its largest, most commercially successful, and strongest sector, the long distance markets. The three billion dollars of public funds that Amtrak invested in the Acela program has earned a negative rate of return (because Amtrak’s NEC and overall financial results subsequent to the inauguration of Acela services have steadily worsened, taking account of the resulting cannibalization of the NEC’s fixed assets to cover overall losses in that segment), yet Amtrak’s only significant new investment proposal is to sink billions more into the Acela successor, in a market chronically afflicted with low utilization, negative returns, and a paltry market share.
Unintended Consequences

Much of what I have related to you is not well understood because few have bothered to look behind Amtrak’s mischaracterizations of its business results. We think Amtrak’s obsessive fixation on the NEC reflects the preferences of its political patrons in congress, management bias, or perhaps its own obsession with the only market where Amtrak owns the railroad it operates over. But the result is the foolishness represented in both management’s endless repetition of the same failed investment strategy, and as reflected in the Administration’s budget proposal. The great irony is that if that budget were to be implemented, according to Amtrak itself, the country would lose the largest, most successful and productive group of trains Amtrak operates, and the annual federal subsidy for what remains, according to Amtrak, would rise not fall. That doesn’t seem like good policy or good business to us.


Amtrak needs to re-think its investment strategies to direct capital to its largest and most commercially successful yet undercapitalized segments. Amtrak needs to come clean with congress, the administration and the public about the financial and operational performance of its three groups of trains using appropriate metrics. The administration could help enormously by demanding that Amtrak measure and report its results using the same metrics that airlines use:  passenger miles, load factor and market share, as well as financial segment results compiled in accordance with Generally Accepted Accounting Principles. Compelling Amtrak to comply with SEC standards for reporting its segment results would also be illuminating and useful, to congress, the executive branch, and the public. The Administration should demand that Amtrak redirect a large part of the capital made available to it, to be invested in its largest and most undercapitalized segment rather than its smallest or second smallest, and most overcapitalized, segment. Only then could Amtrak expect to achieve real growth and real reduction in its endless need for federal subsidy.

I would be pleased to discuss any of this with you if you have any interest in pursuing it.

URPA is an independent institute devoted to research and education on intercity rail passenger issues.
Andrew Selden, President, United Rail Passenger Alliance