Amtrak at 50: Defining the “Basic System” of Service Routes
President Richard Nixon, against the wishes of his senior White House advisers, signed the Rail Passenger Service Act into law on October 30, 1970. (Read the full story here.) The law provided for the establishment of a national passenger rail operating company, informally called “Railpax,” that would take over passenger operations from freight railroads starting on May 1, 1971. It would not own the tracks upon which it operated – Railpax as created would only be a service provider.
But while the government had six months to get the new corporation established, the key question of “which routes will the new corporation serve” wasn’t left up to Railpax. For the first four years, the law required Railpax to serve a “basic system” of endpoint routes to be determined by Secretary of Transportation John Volpe, and it gave him a series of extremely tight deadlines for draw the route map. (After May 1, 1975, the company could start to discontinue unprofitable routes.)
Per title II of the new law, the Secretary was required to submit to Congress and the Interstate Commerce Commission “his preliminary report and recommendations for the basic system” of routes within 30 days of enactment (by November 30, 1970). The ICC and other interested parties had until 30 days after that to review and comment on the preliminary report, and then the Secretary had another 30 days to “give due consideration to such comments and recommendations.”
By 90 days after the date of enactment (January 29, 1971), the Secretary would have to “submit his final report designating the basic system to the Congress,” which would become effective and binding on Railpax immediately upon submission and which was “not reviewable in any court.”
Initial DOT proposal to Nixon.
Even before the law was signed, there was significant back-and-forth between DOT and the White House Office of Management and Budget over the scope of the basic system map. OMB, under new Director George Shultz, had urged President Nixon to veto the Railpax bill because the new railroad “will own passenger trains which are proven losers” and because railroad executives privately forecast “annual operating deficits of $50 million or more” for the new operating company.
With regard to picking the routes, Shultz told Nixon in the same memo that “The problems in identifying a viable grid are monumental. To do it in 30 days appears to risk selecting poor routes and being burdened with an unnecessarily large annual deficit.”
(Unfortunately, the closure of the National Archives from March 2020 to this writing meant that I haven’t yet been able to access the historical record of that back-and-forth. So, thanks, COVID.)
On November 24, 1970, Secretary Volpe sent President Nixon an eight-page memorandum summarizing his recommendations for a basic route map. He wrote that “We have worked closely with the Office of Management and Budget in reviewing the basis for our proposed system as well as to examine potential alternatives.” He then summarized his selected proposal:
My recommendation provides for 27 routes, of which 13 are designated as ‘corridor’ routes and 14 as ‘long haul’ routes. The ‘corridor’ designation applies to a pair of cities each with a population of one million or more and less than 300 miles apart…The Corporation will be required to provide two daily trains in each direction in the corridors and one train in each direction on the long-haul routes. In the higher volume corridors, such as the Northeast Corridor, it is expected that the number of trains operated by the Corporation would be substantially greater.
He explained that “Corridor routes were selected if the demand projected for 1975 was sufficient to permit trains to compete effectively with buses on a time basis and airplanes on a cost basis.” For the long-distance routes, after “a principal requirement that all regions of the country be connected by a reasonably direct route,” DOT looked at existing ridership data and included all existing routes that had a ridership of more than 100 passenger-miles per train-mile (which Volpe said was “a level of ridership at which a train should be approaching a break-even point assuming efficient management”).
(Ed. Note: If that 100 pax-mile per train-mile break-even threshold were ever true, it certainly isn’t true anymore. According to Amtrak, all of its long-distance routes averaged more than 100 passenger-miles per train-mile in fiscal 2019 (pre-COVID), and they all lost money. Excluding the Auto Train, the highest per-mile ridership was the Coast Starlight, which had 199 passenger-miles per train-mile in 2019 and which posted an allocated operating loss of around $150 per passenger.)
Volpe’s memo listed seven assumptions that underlay DOT’s route selection:
- The need to meet the legal requirement of section 201 of RPSA to “take into account the need for expeditious intercity rail passenger service within and between all regions of the continental United States.”
- Preserving “an essential nucleus of rail service throughout the country as an alternative carrier to the bus and airplane.”
- Only selecting routes “where high patronage warrants or where careful analysis indicates a high potential for attracting significant passenger volume in the immediate future.”
- No major capital investment (no “costly improvement in roadbeds”) and “new equipment will be phased in only as demand warrants.”
- Quick profitability and no “need for any Federal operating subsidy other than the $40 million capitalization grant to start the Corporation and the related loan and loan guarantee provisions.”
- Political acceptability: “We believe that the system we designed is the minimum which has a chance to be acceptable to the public and Congress.”
- Being apolitical and using objective analytic techniques to analyze routes and make selections.
Volpe included a map of the proposed routes (they were straight lines, since the basic system plan would only select route endpoints and would leave selections of the specific routes between those endpoints up to Railpax). The crosshatched lines are corridor service and the uncrosshatched lines are long distance service.
In an attempt to show his preferred route map as a middle-of-the-road choice, Volpe’s memo showed three other options considered and rejected – one with eleven additional long-distance routes, one with fewer corridors, and one that dropped all unprofitable routes.
Sec. Volpe’s Basic System Alternatives, As Presented to the President
|Current||DOT 1||DOT 2||DOT 3||DOT 4|
|1975 Projected Profit (M $)||none||$14.1||$24.4||$35.9||$55.2|
|Level of Service|
|Long Haul Routes||25||14||7||6|
|Trains (One Way)||547||135||124||66||45|
|Metro Areas Served||184||145||131||131||66|
|Long Haul Routes||49.2||16.4||13.0||13.5||7.1|
The financial numbers that DOT and Volpe gave to the President showed that all four of those basic systems would reach some level of profitability by 1975. Volpe’s memo even included a projected profit/loss statement for the first five years of Railpax’s operation under the proposed basic system map, in millions of dollars:
|Costs and Expenses|
|Lease of Trains||10.3||10.3||9.5||4.7||4.0|
|Special Train Movement||-1.2||-0.1||1.3||1.6||1.9|
|Income from U.S. Mail||0||3.3||6.7||6.9||7.1|
Some of the numbers in DOT’s financial assumptions were widely regarded, in the industry, as fanciful. For example, the assumption that the Postal Service, after buying a fleet of long-haul trucks and ceasing most transport of mail by rail in the 1960s, would suddenly go back to mail-by-rail in 1972. Among those skeptical of DOT’s numbers was the White House Office of Management and Budget.
OMB responds to Volpe.
Years later, Jim McClellan (one of the principal Federal Railroad Administration staffers tasked with design and implementation of Railpax) wrote of the back-and-forth between OMB and DOT/FRA on the potential economics of Railpax, both before and immediately after the bill was signed:
Another serious conflict arose between the OMB and the DOT/FRA. We had made financial forecasts using a number of models. The Northeast Corridor ridership and economic models, on which millions of dollars had already been spent, were assumed to give the correct outcome for the NEC–wrong! New models were developed for other corridor services and for the long-haul trains, mostly by me…
When all the financial models were cranked up, the result showed a profit for the NEC (all costs in) and losses for most of the other corridors and all the long-haul trains except Chicago-Los Angeles. But the losses were modest, though all profit-and-loss estimates excluded any capital costs (again, grant money was assumed). The projected surplus from the Northeast Corridor would just about offset the losses elsewhere in the system. But it was known that Railpax would never make it as a long-term going concern able to renew its equipment, and the assumption that the government would apply up-front capital proved to be totally mistaken.
The OMB attacked the outcome, as well it should have. The DOT/FRA defended its efforts, knowing that if the financial projections were shown to be wrong, that would be the ballgame: Railpax would go down the tubes, and so too might a lot of freight railroading.
Either the same day as Nixon received Volpe’s memo or the day after, OMB Director George Shultz responded with his own memo to the President framing the subject a bit differently, both substantively and politically. Shultz wanted Nixon to look at four issues:
- “Determination of the basic system of routes;
- “Impact on the labor force;
- “Creation of such an extensive basic system with so many unprofitable routes that the Corporation will not be able to operate on a sound basis; and
- Domination of the corporation by DOT.”
Shultz made the explicit point to Nixon that the whole process was, at least in part, a political exercise designed to keep the Administration from being blamed for the collapse of passenger rail service (which was why Nixon signed the bill in the first place), and that also should make a difference in the route network selection. As far as Shultz was concerned, “Profitability and financial soundness of the system are mandatory so that your Administration cannot be blamed for the failure of the Corporation, the demise of intercity rail passenger service, and possibly the ultimate nationalization of the railroads.”
According to Shultz, this didn’t just extend to the route selection – since Shultz felt from the beginning that there were “very substantial risks” of those harms, “the Secretary and the Department of Transportation would have to minimize their involvement in the Corporation.” Instead, Nixon should appoint “a board of strong, independent, profit-oriented individuals” to try and make the railroad succeed (and take as much blame as possible if it failed).
Shultz presented his own range of options:
- Profit-maximizing (every route must be projected to be profitable by 1975), which only meant six routes: the Northeast Corridor, NYC-Buffalo, DC-Miami, Chicago-Miami, Chicago-New Orleans, and Chicago-L.A.
- All the routes in option #1 plus routes with projected 1975 losses of less than $1 million per year (Chicago-Minneapolis-Seattle, Kansas City-L.A., DC-New Orleans, Philadelphia-Pittsburgh-Cleveland-Chicago, Detroit-Cincinnati, and Cincinnati-Cleveland).
- The DOT proposal with train frequency cut back to minimize losses (cutting down to one train per day per route everywhere except the NEC and NYC-Buffalo, and cutting service on the Seattle-San Diego and Houston-L.A. routes to just three trains per week).
- Option #3 plus upgrading one route each for the Midwest and West to more frequent “corridor” service (Cincinnati-Cleveland and L.A.-San Diego).
- The DOT proposal, presented as the maximum level of service that should be considered (“I see no need to consider seriously anything more extensive than the DOT proposal for the initial announcement of the basic system”).
The route maps for the five options can be seen full-size in the actual document but are reproduced in miniature below.
Also, while Volpe had stated that DOT believed that the 17,000 railroad employee layoffs that would take place under his basic system “can be handled in an orderly manner without labor unrest” (albeit at a cost to the railroads of as much as $300 million in termination costs), Shultz said that the probability of a crippling nationwide railroad strike was difficult to assess and should not prejudice plan selection: “Would the difference between a 60 percent reduction in employment (option #5, DOT’s) not cause labor unrest whereas an 80 percent reduction (option #2) would?”
The OMB memo raised the thorny issue of the degree to which rail riders in the Northeast should subsidize the rest of the system: “
Shultz presented OMB’s own matrix of pros and cons of each of the five routes, starting with OMB’s own estimation of Railpax profit/loss in 1975 under each plan. While Volpe and DOT had predicted a $24 million profit in 1975 under their route map, OMB thought that system would really come in between a $12 million profit and an $11 million loss.
OMB’s matrix also contrasted the buy-in cost and severance pay costs for the railroads, the percent of population served and train-miles operated for each option, and ended with something that Nixon’s political side would, no doubt, focus on: Shultz’s estimate of “Presidential risk of blame for killing RR passenger service,” both short run (before 1973, i.e. in time for Nixon’s 1972 re-election campaign), or in the long run.
The copy of the Shultz memo we found in the Nixon Library (in Bud Krogh’s files) was pre-signature, so we don’t know which of the five options Shultz recommended. We think, based on context, it was either #2 or #3 (someone, we’re not sure who, circled option #3 on the file copy).
But with regards to avoiding blame wherever possible, Shultz recommended that Nixon order Volpe to “Make no distinction whatsoever between ‘corridor service’ (higher frequencies) and ‘long haul’ (one daily or less), and to “Avoid specification of frequency of service.” The law did not require the basic system map to specify frequency, and vagueness in the beginning could buy flexibility in choosing service frequency later on.
Nixon’s initial decision.
On November 25, Shultz sent Volpe a memo on Nixon’s behalf informing him that “the President has decided to use an approach somewhat different from that reflected in the Department of Transportation proposal.”
Nixon selected OMB’s option #2, the one which only had routes which were projected to be profitable, or only lose $1 million per year or less, by 1975. This would have been only thirteen end-to-end corridors:
|Chicago-New Orleans||Chicago-Los Angeles||Chicago-Seattle|
The OMB letter also ordered Volpe to obey Shultz’s suggestions as to vagueness (don’t talk mention train frequency, don’t differentiate corridor service from long haul service, don’t specify any points served between route endpoints, and don’t talk about which, if any, new equipment will be purchased).
There was some additional back-and-forth between DOT and the White House over the next three days, and three additional endpoint routes were added to the preliminary basic map: Chicago to San Francisco, Chicago to St. Louis, and St. Louis to Washington, D.C. (Chicago to Cincinnati was also substituted for Detroit to Cincinnati.)
Secretary Volpe released his “Preliminary Report on Basic National Rail Passenger System” on November 30, 1970 (the statutory deadline). Per his instructions from OMB, Volpe did not make any distinction between corridor trains and long-distance trains, instead stating that “Trains will be scheduled to serve markets at reasonable hours, consistent with demand” and that “The scheduled frequency of trains shall be one or more per day in each direction between endpoints keeping with demand, unless the patronage and other factors relating to a particular route clearly indicate that a frequency as often as one per day in each direction is not warranted.”
The preliminary report also listed potential lines and stops for each endpoint pair on the basic system, but said it was “giving the Corporation the necessary flexibility to choose the line or combination of lines which will best meet the needs of the market in terms of supply/demand, service, and profitability.”
There were several instances of where competing railroads had, over the years, built competing lines between the same endpoint. If Railpax took everything over, maintaining lines to compete with itself would be counter-productive. For example, here is how the preliminary report showed the choices that would face Railpax to select the Chicago-Miami line, depending on what combination of Illinois Central, L&N, Penn Central, Southern Coast Lines, and Central of Georgia right-of-way they wanted to use:
Under the law, the preliminary report was then subject to comments from the public and review by the Interstate Commerce Commission.
The public complains, the ICC reviews.
Volpe’s announcement of the preliminary plan made front-page headlines across the country, with many of the headlines focusing on the number 14 – the number of cities represented in al the endpoints, which didn’t sound like a lot. Someone at DOT was quick to leak to the Washington Star that DOT had wanted to keep DC-Chicago service and New Orleans-Los Angeles service, but that the White House had killed those routes.
Within 48 hours, the pushback was coalescing, beginning on the West Coast, where members of Congress had begun uniting in a call for continued north-south service up and down the Coast. In particular, Sen. Warren Magnuson (D-WA), chairman of the Senate Commerce Committee, criticized the lack of West Coast service on December 1. He then joined the other Senators from WA-OR-CA in writing to Volpe to complain, as did 37 members of California’s House delegation.
A December 6 New York Times headline said “Protests Made Over Gaps in Railroad Passenger Plan” and cited the lack of continued service to Tampa and St. Petersburg. A bill was introduced in the lame-duck session of Congress by Senators from Minnesota and Wisconsin to postpone Railpax’s startup date by six months to give Congress more time to consider the route map. (S. 4574, 91st Congress).
Beyond the strictly parochial complaints, Robert Gallamore and John Meyer (in American Railroads) later wrote that there were two more principled critiques of the preliminary map. One was from the efficiency types who “argued that Amtrak’s comparative advantage lay in medium-distance corridors (such as New York to Buffalo) and that transcontinental service was a waste of time.” The other group believed in the central importance of a national network, and by that measure, leaving so much of the U.S. without service was a significant error.
The Interstate Commerce Commission, which had been analyzing and regulating U.S. passenger rail service for 80 years by that point, issued its own review of the DOT preliminary system outline on December 29. Under the imprimatur of chairman George M. Stafford, the Commission criticized DOT’s decision only to select endpoints, and not the cities and lines served between the endpoints, because that made it impossible to tell if the preliminary system fulfilled the law’s intent to serve as many cities as possible: “As we understand the statute, the intended description of ‘service characteristics’ calls for a degree of specificity not reached in your preliminary report.”
With regards to specific cities served as endpoints and points between, the ICC made six “essential recommendations” and nine “secondary recommendations.” On the essential side, within the endpoints already selected in the preliminary report, the ICC recommended that the Chicago-San Francisco route definitively include both Denver and Salt Lake City and that the Chicago-Seattle route include Minneapolis-St. Paul. The ICC also recommended new endpoint service between Seattle and San Diego (for reasons that included alleviation of highway congestion) and between New Orleans and Los Angeles (to link the South and Southwest to the West). And the ICC recommended that direct through-service between Washington DC and Chicago be retained, and that Tampa-St. Petersburg be reached by service somehow.
On the secondary list, the ICC’s recommendation list included traditional big cities (a Kansas City-St. Louis line, a St. Louis-Detroit line, and a Boston-Chicago line that went more directly than through New York City), some less obvious routes (Salt Lake City-Butte, Norfolk-Charlottesville, and some international connections (Seattle to Vancouver BC, Chicago-Toronto, NYC-Montreal, and Chicago-San Antonio-Mexico City).
The ICC report, naturally, included its own proposed map of these routes.
The ICC recommendations provided excellent political ammunition for the Members of Congress from the affected areas to use in fighting the preliminary map. The problem was, Congress wasn’t around. The lame-duck 1970 session had run so late that, to give everyone a rest, the outgoing 91st Congress provided by law that the new 92nd Congress would not take office until January 21, 1971. And in those days before cable TV news, that meant that there were no major news stories emanating from Capitol Hill until after the 21st.
On January 22, OMB Assistant Director Donald Rice send Director Shultz an options memo defining three possible sets of changes to the November 30 preliminary system:
- Option A. The November 30 route structure “plus the following specified clarifications, which have no incremental financial effect because they were intended in the original structure:” Washington-Chicago, NYC-St. Louis, and including Tampa as a stop in both the NYC-Miami and Chicago-Miami routes.
- Option B. The November 30 route structure plus three more routes, which the OMB staff felt were “the maximum necessary response to the Railpax Act and to public reaction generated by the November 30 announcement:” St. Louis-Kansas City, L.A.-Seattle, and New Orleans-L.A. (the last one only triweekly). DOT estimated that those additional routes would mean $2.6 million in extra losses in 1975; OMB estimated $4.8 billion in extra losses.
- Option C. The DOT proposal, which would meet the “essential” ICC recommendations. It included Option B (the Nov. 30 prelim plan plus three more routes), plus Cincinnati-Norfolk, Salt Lake City-Portland, Boston-Chicago (triweekly), and L.A.-San Diego. It also made two international routes, Seattle-Vancouver and NYC-Montreal, “recommended” (not mandatory) routes. Total 1975 cost was estimated by DOT to be an additional $3.1 million in losses in 1975 (in addition to the Option B losses), and estimated by OMB to be an additional $5.2 million.
Because of the closure of the Archives, we haven’t yet been able to get the full paper trail on the final decision. But DOT and the White House wound up with a modified Option C, rejecting the Salt Lake City-Portland and Boston-Chicago routes while including Cincinnati-Norfolk and the extension of the north-south West Coast route from Los Angeles to San Diego (as well as the two international “recommended service” options).
Volpe released the Final Report on the Basic National Rail Passenger System at a press conference on January 28, 1971. In his remarks, Volpe said that since the preliminary system’s release on November 30, “I have been one of the most communicated with – or at – men in the country. And I freely admit that all of my mail has not been fan mail.”
As he announced the five additional city-pairs (DC-Chicago, NYC-Kansas City, San Diego-Seattle, L.A.-New Orleans, and Norfolk/Newport News-Cincinnati), and the announcement that “it was the intent of the initial report” that Tampa/St. Petersburg be served all along, Volpe mixed a statement of hope with a warning:
It is my hope that all points designated in the Basic System will provide the basis for a financially viable intercity rail passenger system.
By providing improved service, clean and reliable equipment and more efficient utilization of facilities, the Corporation has the opportunity to reverse the long-time decline in rail ridership.
If, in fact, the traveling public does not respond to the Corporation’s efforts by increasing ridership, the Corporation then has the option of dropping unprofitable routes in the interest of preserving the viability of the total passenger system.
The incorporators appointed by President Nixon would have just three months to select the lines to serve within (and possibly beyond) those basic system endpoints, negotiate contracts with the freight railroads for their Railpax buy-in contributions and their terms of service, and do all of the other overhead work necessary to get the new nationwide passenger railroad operating company up and running. The starting date set by the RPSA law was May 1, 1971.
To be continued…
 Memorandum from OMB Director George Shultz to President Nixon, undated, subject line “Railpax.” Original located in White House Central Files: Subject Files, Series TN 4 (Railroads), Box 6, Folder “5/1/70 – 11/5/70,” Nixon Library.
 Memorandum from Transportation Secretary John Volpe to President Nixon, November 24, 1970, subject line “Designation of the Preliminary Routes to be Served by the National Railroad Passenger Corporation.” Original located in materials of Egil “Bud”Krogh (White House Special Files: Staff Member and Office Files), box 36, folder entitled “Railpax 1970,” Nixon Library.
 Jim McClellan, My Life with Trains: Memoir of a Railroader. (Bloomington: Indiana University Press, 2017) pp. 190-191.
 Undated memorandum from OMB Director George Shultz to President Nixon (but it has to be November 24 or 25, 1970), subject line “Railpax issues for your decision.” Original located in materials of Egil “Bud”Krogh (White House Special Files: Staff Member and Office Files), box 36, folder entitled “Railpax 1970,” Nixon Library.
 Memorandum from OMB Director George P. Shultz to Transportation Secretary John Volpe, November 25, 1970, subject line “Railpax.” Original located in materials of Egil “Bud”Krogh (White House Special Files: Staff Member and Office Files), box 36, folder entitled “Railpax 1970,” Nixon Library.
 U.S. Department of Transportation. Preliminary Report on Basic National Rail Passenger System. November 1970.
 Vera Hirschberg, “Transportation Report/Critics pressure Secretary Volpe for bigger rail passenger network,” National Journal, January 16, 1971 p. 128.
 Robert Gallamore and John R. Meyer, American Railroads: Decline and Renaissance in the Twentieth Century (Cambridge: Harvard U. Press, 2014) chapter 11.
 United States. Interstate Commerce Commission. Review of Preliminary Report on Basic National Rail Passenger System, December 29, 1970 p. 7.
 Memo from Assistant Director Donald B. Rice to Director George Shultz, January 22, 1971, subject line “Final Route Structure for Railpax.” Original located in materials of Egil “Bud”Krogh (White House Special Files: Staff Member and Office Files), box 36, folder entitled “Railpax-Amtrak 1970-1972,” Nixon Library.
 John A. Volpe, “Statement of Secretary of Transportation John A. Volpe At A News Conference, Thursday, January 28, 1971, Washington D.C” (Attached to DOT press release number 2071 of that same date.)