The current round of maglev studies is just the latest of a series of high speed train studies in North America that have extended over more than 20 years. The history of passenger rail service in North America is a tremendously complex story in itself and we won't try to do it justice here. In the U.S., a country that was transformed by the construction of a massive rail network, that network was essentially supplanted by an equally ambitious and formative highway network. Passenger rail has been diminished to a shadow of its former self. In Canada, the decline of passenger rail service has not been as precipitous but has nonetheless been significant.

In the 1970s and 1980s, many Americans experienced the "Bullet Train" in Japan and the TGV trains in France and wondered why we weren't considering something similar over here. Studies were initiated in several regions to determine the costs and potential market for high speed train service. The earliest round of work was conducted by rail advocates who, many feel, overstated the market potential and understated the costs of this service. As a result of this criticism, the High Speed Rail Association adopted guidelines for the estimation of markets and costs.

Over the past 20 years Resource Systems Group, SurveyCafe's proprietor, has been contracted to conduct rigorous market research for a series of high speed train studies, spanning the North American continent. The work requires much more sophisticated approaches than are used for most other market research. This is because:

1. The level of investment that is required to build these systems is in the billions of dollars. Revenues from passenger fares will need to cover at least part of this investment (plus the operating costs), so the market estimates need to be "investment grade" - rigorous enough to withstand detailed scrutiny by those who would invest huge amounts of money in the system. 2. The door-to-door travel times that are provided vary significantly depending on where a traveler is coming from and going to. Trip locations are scattered widely in most U.S. corridors. This means that there is practically infinite variation in the competitive conditions between the train and other types of transportation. 3. Travel alternatives are not generally "perfect substitutes" for one another. For example, car trips can become too long to be practical for a one-day business meeting but these same trips could be easily made by much faster trains and planes. This means that travel market forecasting models must relax the simplifying assumptions most commonly used in market research. 4. New service will "induce" demand - Major improvements in transportation stimulate demand and the forecasting models must be able to estimate the level of induced demand as a function of the transportation service levels.

Our work analyzing markets for high speed trains began while we were professors and researchers at Dartmouth College. Our first study was for the Pennsylvania High Speed Rail Commission, which was evaluating high speed train service between Philadelphia and Pittsburgh. We then began a series of three studies of the Tampa to Orlando to Miami market. These started with work for a consortium who were evaluating the feasibility of privately-financed service, continued with a large Florida DOT-funded study of the statewide market for high speed trains and concluded most recently with detailed market research for a public/private partnership that had been awarded the "concession" for a 180 mph Tampa to Orlando to Miami service. We have since conducted detailed market research for proposed services in the corridors of Cleveland to Columbus to Cincinnati, St. Louis to Chicago to Milwaukee, Toronto to Ottawa to Montreal to Quebec City and Seattle to Portland. High speed train service has also been evaluated in Texas, California (www.cahighspeedrail.ca.gov), New York the Southeastern U.S. (www.sehsr.org) and numerous other regions.

back to top

This, of course, is the multi-billion dollar question in North America. Early studies pointed at the experience in Europe and Japan and extrapolated that ridership to the U.S. That approach was widely criticized because of dissimilarities in land development patterns and in U.S. travelers' behavior. Highways have allowed U.S. development to become dispersed away from the traditional urban core areas. U.S. travelers are used to driving - some would say we are wedded to our cars - and many have ridden neither a train nor any other form of public transportation.

The work to determine how many would ride trains has taken many forms. Each market is quite distinct along many dimensions (e.g., number of travelers, type of travelers, competitive conditions) and this has meant that separate studies are required for each region. Some of the studies have relied on data adapted from other U.S. market research, but most have included region-specific market surveys. In most cases, the surveys have included questions to establish the conditions under which travelers of different types would use high speed trains.

We have used traditional focus groups to understand the general feelings about high speed trains among American travelers and, in the case of the recent Florida work, to develop initial marketing strategies.

Our high speed train surveys have used computer-based questionnaires that collect detailed information about actual trips that are made by respondents. We use maps and databases within the software to determine precise trip origins and destinations. We build simulation models to construct realistic high speed train alternatives, customized to the respondents' trips. And, we use advanced statistical methods to analyze patterns of response in order to develop models that predict ridership levels as a function of train fares, speeds and the costs and travel times of competing modes such as auto and air. These models, in turn, are used in detailed "network forecasting" exercises to determine the share of each travel market that will be attracted to the new service. "Detailed" means analyzing up to 1 million geographically-distinct trip origins and destinations, each for up to a dozen traveler types.

So what have we learned about the market for high speed trains from all of this work? First, while we may be wedded to our cars, this is at most a marriage of convenience. That's the good news. The bad news is that the convenience we demand will not be easy for high speed rail to provide for much of our travel. High speed trains could be successful in the relatively limited number of North American corridors which have a high density of travel between concentrated centers of activity. High speed trains generally compete most effectively against air and car for intercity trips of 100 to 600 miles in length; the proposed maglev initial segments are all less than 100 miles but represent unique markets that would be subsumed in a much longer system.

However, it's the details of how a service is provided that will allow it to succeed in these markets. The service will, and should, look very different across the broad spectrum of geographic and traveler markets that it serves. In each segment, however, it must provide convenience and value relative to the alternatives and it must do so reliably. What form will the service need to assume to accomplish this?

Convenience - There is a guide to U.S. airports that lists travel times from the airports to downtown by public transit and by car (or taxi). In many cases, the listed public transit times are less than or equal to the car times. Yet, the car access volumes are tenfold (or more) higher than the transit volumes. Why? First, because most trips to and from U.S. airports are not to the downtown area and, second, because the quoted public transit travel times are just the riding times and ignore the access and wait times. We know from our research that travelers consider their full door-to-door travel times, and in many cases they treat the waiting and access times as much more onerous than the riding times. Where car is the benchmark, convenience means a lower total travel time and the flexibility to make the trip at any time of day. This can be accomplished only if good connecting service is provided at the train stations. However, there are also ways to reduce the "onerousness" of the wait and access times: provide in-station amenities that allow the waiting time to be productive and/or pleasant and make the transfers between travel modes as seamless as possible. Where air is the benchmark, trains can compete effectively by requiring less time for check-in, passenger screening and boarding. They can also provide at least equivalent on-board amenities to make the riding time productive. However, the train stations have to be able to compete with the many amenities provided by modern airports: a wide choice of access modes including free shuttles, supershuttles, taxis, rental cars and the ubiquitous car parking offered with tiered pricing (less expensive remote lots, more expensive in-close short-term spaces); curbside pickup/dropoff/baggage checking; a wide selection of restaurants and shops; and lots of places to get travel information. Value - There is the equivalent of a holy grail in intercity travel - totally price-insensitive business travelers for whom value is a function only of the quality of service provided. Such travelers do exist as can be evidenced by those who pay $3,000-$5,000 more for the slightly wider seat and (much) better meal provided in business or first class on a 6-hour transatlantic flight. On the other hand, the success of low-price airlines testifies to the existence of a segment for which price is a key consideration. The market for high speed trains spans this spectrum. The question for high speed rail is what the value proposition should be across its many markets. The answer, we believe, includes general positioning as a "premium" service and selective use of airline-style differential service classes, yield management pricing, and loyalty programs. Trains will find it difficult to compete toe-to-toe on price with both the low-price airlines which enjoy free airspace (as compared with the cost of constructing and owning tracks) and with cars, which enjoy "freeways" and which can, in effect, accommodate parties of up to five passengers for the price of one. Reliability - American travelers are not currently served by any truly reliable intercity travel options. Major airports are at capacity, making the entire system highly vulnerable to delays of all types. Highways in and around major urban areas are chronically congested. And current passenger rail service is adversely affected by a number of factors including sharing of tracks with slow freight trains and inadequately maintained equipment. High speed trains with modern equipment, running on dedicated tracks should be able to equal the European standard for reliability - you should be able to set your watch by when the trains arrive and depart. Our research indicates that many travelers will pay a significant premium for this level of reliability, assuming that the other elements of value and convenience are adequate. Market responsiveness - The concern that many have about high speed trains is that they require infrastructure that, once placed, cannot be moved. It will take 20-40 years to amortize the costs of that infrastructure and, in the meantime, the market may change radically, turning the high speed train that met the needs of 2010 into a dinosaur by 2030. For example, surely the Internet, videoconferencing and other technology advances will reduce the need to travel. We can't rule that possibility out, but history indicates that radical advances like the telephone in fact stimulated travel by allowing people to remain connected across greater distances. That connectedness creates a demand for a type personal contact that can best be accomplished by traveling to meet "in person". Much more likely is that the types of demand for travel will shift. Some of the shifts could well be favorable to the service - research indicates that major transportation can stimulate demand over the longer term by encouraging complementary development patterns. High speed train service will need to be able to redefine itself to anticipate and respond to those and other shifts. This means that the organization that runs the service must be nimble and highly market-focused.

So, assuming there is a service that meets these basic conditions, will enough people ride high speed trains to justify their construction in the U.S.? Again, it appears likely that only a selected set of corridors will be able to support such a service with sufficient ridership. The projected ridership levels for the proposed Florida system were sufficient to cover costs with some level of public involvement. The results from the maglev studies are not yet public but it appears likely that some of the proposed plans will be viable given $1 billion of federal construction funds.

Our work in this subject area is ongoing, so we would especially appreciate your comments about how high speed trains in North America might serve your travel needs.

back to top