Missouri High Speed Rail: Mission Impossible? | Part II

Tweet about this on TwitterShare on Facebook0Share on Reddit0Share on LinkedIn0Print this pageEmail this to someone

In part one of this three-part series we looked at the state’s infrastructure and decided to build a high speed rail line between Kansas City and St. Louis. Let’s continue.

Now that we have funded the track, let’s talk about operating the trains: Could a high speed train be operated and maintained profitably? There are two possibilities: Amtrak could buy and operate HSR trains, or the State of Missouri could issue an RFP for a private company to operate the trains on their track. This model is already being used in Europe and Asia. Private operators pay a user fee to operate their trains on government-built tracks.

But for now, let’s put Amtrak in charge and do a little math.

{Concept Amtrak River Runner high speed train on dedicated track}

The state of Missouri would grant Amtrak the usage of their high speed track at the cost of maintenance; initially $100,000 per mile per year, which would add up to a $25 million annual user fee for the entire track.

[This is half the European average of ~$200,000 per mile per year, but the tracks there are being much more heavily used with up to 15 trains an hour.]

Missouri has now secured the upkeep of the track for the duration of the contract – say, 30 years – at no cost to the tax payer. [Compare this to a highway.]

ScreenShot public-private model

In the northeast corridor, Amtrak has a ridership of 11.5 million per year in a catchment area of 44 million. 1 out of 4 takes the train once a year. By using the same ratio for our KC-STL line – with its catchment area of 5 million – we project 1.25 million yearly or 3,400 daily passengers; enough to warrant 12 one-way or 6 roundtrips. Such a schedule could be operated by two (400-seat) train sets, each making 3 round-trips.

The estimated* operating costs for our 248-mile trip is projected at $0.20 per passenger mile or $49 per passenger, including acquiring the trains and the cost of personnel and maintenance, based on a 70% load factor. The $25 million per year user fee for the high speed track translates to another $0.08, for a total of $0.28 per passenger mile, or ~$69 per passenger for our 248-mile trip.

[*Amtrak 2011 cost per passenger mile on the Chicago-St. Louis line was $0.17 and on St. Louis-Kansas City $0.13 with load factors just under 50%. The lower the load factor, the higher the cost per passenger mile and vice versa.]

A review of ticket prices on the Amsterdam-Paris Thalys line shows that business class tickets are priced at $45-$85 and first class tickets at $100-$125 for a 310 mile trip. For our 248-mile KC to STL line this would amount to $36-$68 for a one-way business class ticket and $80-$100 for a first class ticket. Using a 2:1 ratio business/first class tickets, and a 2:1 ratio walk-up/early rate, the average ticket price would calculate to $69 which exactly matches our initial operating cost estimate of $69 per passenger.

What this means is that based on the projected initial ridership the trains could be operated at the break-even point with ticket prices at a level comparable to that in Europe. Once ridership exceeds the initial projections, operating the line would become profitable.

An initial daily schedule could look like this:

Schedule - web

The next question we’ll have to answer is: Can HSR compete with airplanes and cars?

We’ve already established that highways don’t pay for themselves. Now let’s take a look at commercial aviation. The Airport and Airway Trust Fund supporting the nation’s aviation system currently receives about 35% of its funding from the U.S. Treasury. [It also faces a yearly shortfall of about $16 billion for the foreseeable future.]

Despite the airline system being hugely subsidized, you’ve probably heard this perennial joke: “You want to become a millionaire? Start an airline as a billionaire.” Airlines don’t make money. They make money when times are good; they lose money when times are bad. Every major U.S. airline (except Southwest) has filed for bankruptcy over the last decade; some more than once, and many have not survived.

Going back to Missouri, let’s compare the time and cost of making the trip from downtown St. Louis to downtown Kansas City by different modes of transportation:

Airplane:15-minute drive to the airport; at airport one hour before flight; one hour flight, gate-to-gate; 45 minutes from gate to downtown by taxi. Ticket prices are based on the cheapest non-stop air fare and walk-up fares found on Southwest.com. [We did not consider a ~$50 one-way taxi fare from airport to downtown.]

Car:Trip from home to downtown at maximum highway speed. Cost based on IRS standard of $0.50 per mile, including gas and wear.

HSR train:A 15-minute drive/walk to the train station; at station 15 minutes before departure. Ticket prices are based on operational cost per passenger mile.

The conclusion has to be that an HSR train between Kansas City and St. Louis would be competitive. It would be the fastest- and by far the most economical option.

Based on the initial projected ridership it would also remove some 1,400 cars per day from I-70; the bulk of that number during rush hours.

[Based on 2/3 of HSR passengers being former drivers: 3400 X 2/3 ÷ 1.6 occupants per car = 1,416 cars. Current car count on I-70 between KC and STL is 30-35,000 per day.]

In Japan (HSR) trains capture 40-50% of all passengers in the 188-469 mile range.
{In Japan (HSR) trains capture 40-50% of all passengers in the 188-469 mile range.}

The 248-mile distance between Kansas City and St. Louis happens to be in the sweet spot for high speed rail. This is the range wherein fast trains beat out both airplanes and cars.

In part three we will attempt to answer what the economic, environmental, and safety advantages would be, and how many jobs would be created by HSR in Missouri.

Tweet about this on TwitterShare on Facebook0Share on Reddit0Share on LinkedIn0Print this pageEmail this to someone
  • jmirowitz

    High speed rail is so 2000 and late. Elon Musk’s vision for Hyperloop would be a much better solution!

  • dempster holland

    Several points:
    1. What was the KC/Stl traffic in the peak year of train travel, whatever
    that peak year was (exclude WW 2 years). Then adjust by deducting persons
    transfering from other segements (eg, Indianapolis-st louis)
    2. Obvioiusly, right of way costs could be less if part of the existing
    row for UP tracks could be used. Also, what are the cost savings, patronage
    impacts for a slightly slower tain?
    3. Metro has said it would consider rail commuter if high speed rail was
    implemented, sharing an existing or a widened existing row.
    4. The route through kirkwood has the advantage of being better
    combined with local commuter than the airport route, but the airport route
    of course could replace some of the small commuter air flights.
    5. A possible source of funds would be to stop increasing airport
    capacity and use airplane fees to increase highspeed rail capacity

    • Eric

      The airport station would be a complete waste. There are some place where an airport station makes a lot of sense, like Chicago or Atlanta. Most flyers have to switch planes there anyway, so why not make one of the two flights (the shorter one) into an HSR trip. But nobody switches planes in STL anymore, and if you’re only making one train/plane trip, better to have it end in the most useful location for the average traveler, i.e. downtown. What makes the idea even worse is that the rail line on the opposite side of the airport from the terminal! Even if you wanted to transfer from train to plane, there would be no way to do it.

      Now, it’s not a terrible idea to put a second station west of Downtown, so people in West County don’t have to drive all the way east and then take the train west. But this station should be in Wentzville or St Peters (or Kirkwood), not at the airport.

  • Donald

    The assumptions being used are quite the stretch, if not outright wrong. 1 in 4 KC and STL residents will use the train a year? HAHA, yeah right. You cannot compare the ridership of NY-Boston-DC to STL-KC. Residents in the Northeast actually have reason to go to each of those cities on a pretty regular basis, not to mention the number of people riding the train does not mean only residents. Millions of tourists are riding those trains, as well. Last time I checked, STL and KC were not tourist hot spots, especially not for the train riding type tourists.

    How many flights are between KC and STL daily? 6? 8? So, you are looking at maybe 600-1000 passengers a day who fly between the two cities (and you have to assume that all those passengers are only going to KC and are not only making a connection, which is again not true).

    Get your assumptions at least a little bit in line with reality, then come make a case to the public.

    • Alex Ihnen

      I think you misinterpret the numbers a bit. Yes, many assumptions are made (just like when a highway is built, or a lane added, or whatever), but the 1 in 4 number doesn’t really mean 25% of all residents will use the train. The northeast numbers mean that 11.5M trips are taken on the train in an area with 44M residents. As you mention, not all those trips are by residents and more importantly, many of those trips are taken by the same people. One person riding 20x, for instance, equals 20 of those trips. But yes, assumptions are very tricky.

      High-speed rail hold the opportunity to make STL and KC one tourist destination. It would make a great vacation to arrive in one city, tour, take the 2hr HSR trip to the other, and explore some more. The draw of one city might not cut it for everyone, but combine the two and add in a convenient and fast trip in between and the equation changes. I’d agree that if everything else stayed the same, an HSR line in Missouri might not be a great investment, but the point is (at least to me), that such an investment changes the playing field.

    • FrankDeGraaf

      Currently, 600 daily passengers already ride the train between KC and STL, (an increase of 68% since 2007, despite the fact that there are only two trains daily at a trip length of 5 hours and 40 minutes).

      HSR would likely capture 65% of the current ~900 daily flyers = ~600. A number of Southwest chief passengers would switch to KC-STL-CHI. Let’s say an additional 200. That’s 1,400 daily passengers.

      Add to that 1,000 daily drivers each way (=1,400 cars out of a total of 30,000-35,000 daily cars between the two cities) and there are your 3,400 daily riders.

      I have not taken into account the increase in passengers when Chicago-STL 110 MPH comes on line. CHI-STL ridership has increased 217% since 2004. The frequency of the CHI-STL line will go up to 8 daily trains, maybe even more.

      I believe these assumptions are in line with reality and even on the low side if you look at the much larger increases in ridership on every new HSR line worldwide.

  • Daniel Layton

    I’ve lived in St. Louis for 5 years and have never had a reason to go to Kansas City. When I lived in D.C., I had to go to New York or Philadelphia or Baltimore on a rather regular basis for various reasons. Obviously my personal experience won’t extrapolate to everyone, but a rail line between STL and KC seems intuitively much less useful than the NE Corridor, which would mean much lower usage. The NE Corridor includes like 10 big cities with large populations. 1 out of 4 take the train once a year because there are many destinations to choose from. That is to say, Amtrak gets a share of the D.C. to NY traffic, the BMore to Philly traffic, the Wilmington to Boston traffic, the New Haven to D.C. traffic, etc. etc. This line would basically only get a share of travel to STL, KC, Jeff City, and Columbia, all of which would almost certainly have lower per capita demand anyway than anything in the NE. This is a big problem for ridership. Do we know how many people travel by rail + plane between KC and STL daily? This would be a better metric of determining usage volume than a fraught comparison to the NE, which is a totally different beast.

    • Larry Guinn

      Columbia and Jefferson City are destinations to themselves, at the least for state government and universities. STL and KC riders would also be going there and vice-versa. You have to consider the vehicle traffic on I-70, I-44, Hwy 50, Hwy 54 with a percentage of those drivers choosing rail and current Amtrak riders.

  • Taylor Haywood

    There’s one crucial assumption which I find problematic: that the ridership ratio within the catchment area would match that along the Northeast Corridor. The two regions are simply not the same. I lived in St. Louis for nine years, but moved one year ago to Providence, Rhode Island, which is served by all Acela and regional Amtrak trains traveling between Boston and New York.

    The basic issues are 1) that people in St. Louis and Kansas City are overwhelmingly more likely to own cars, and 2) that visitors to either city are more likely to require the use of a car during their visit. There is, therefore, a strong incentive for St. Louis-Kansas City travelers to drive.

    When I lived in St. Louis, even though I walked to work and only used my car a few times a month, I still considered it necessary to own one. (When I moved to Providence, I went car-free.) In St. Louis or Kansas City, particularly as a visitor, public transit would not be sufficient for convenient for me, and the additional cost of taxicabs or rental cars would easily outweigh any savings.

    On the other hand, in the major cities along the northeast corridor (New York, Boston, Philadelphia, and Washington DC) a car is almost always unnecessary, if not actually burdensome. I travel frequently to these cities, and almost always using Amtrak. My friends in Providence who own cars also typically choose Amtrak when traveling to the larger cities in the region because it is not convenient cost effective to take a car into a large city.

    Wikipedia ranks the 50 US cities with the highest percentage of households not owning any cars. Nine out of the top ten of these cities are served by the Northeast Corridor. St. Louis ranks number 22, but Kansas City doesn’t even make the list.

    Secondly, think it is very likely that a larger percentage of the people living along the Northeast Corridor travel regularly between those cities, compared to the St. Louis-Kansas City route. Here I can only speak from my own experience, but in my nine years in St. Louis, I made more than 20 trips to New York, Boston, Philadelphia and/or Washington, and only one trip to Kansas City.

    All that being said, I love the idea of more and better rail service
    in the St. Louis region, and I hope you’ll continue to examine it
    in-depth. My comments are in the spirit of constructive contribution.

    It would be better, it seems, to compare the number of trips taken between the cities in question. Since data for trips by car would be hard to come by, and since there are very different incentives for and against driving in these regions, I might suggest comparing the number of passengers traveling by air between pairs of Northeast Corridor cities (with a similar flying times as KCI-STL) with the number of Amtrak passengers between those same pairs. That way you’re dealing only with people who would be arriving without cars anyway, and you’re also excluding the short-distance Amtrak passengers, such as people who use it for daily commutes (Providence-Boston, for instance).

    • FrankDeGraaf

      Taylor, thanks for your comments. All good points. With regards to vehicle ownership: vehicles per household in Northeast = 1.58; in Midwest = 1.95. So, yes, more vehicles here, but I wouldn’t call it overwhelmingly more.

      With regards to your point about travelling to KC: When the French decided to build the TGV from Paris to Lyon, many made a similar point: “Who is going to travel to the southern, somewhat sleepy town of Lyon?” The cities of Paris and Lyon had no correlation. Now it’s the most profitable high speed train route in the world.

      Keep in mind that when Chicago-St.Louis (medium speed) comes on line this will also benefit the STL-KC line. Travelers from KC could take the 5:30 AM train to STL and then the 8:00 to Chicago, and get there by noon.

      I will address some of the points you made in part three. But agreed, there is a lot more research to be done. The article is certainly no scientific paper and merely meant as a conversation starter.

      • Eric

        “Travelers from KC could take the 5:30 AM train to STL and then the 8:00 to Chicago, and get there by noon.”
        A 6.5 hour rail trip from KC to Chicago? That will certainly be competitive with the 1.5 hour flight (3.5 hours if you include security, delays, time to reach airport, etc.)

        • FrankDeGraaf

          You are correct. It won’t compete with air travel, but train riders currently on the Southwest Chief would switch to the route over STL. Once Chicago-St. Louis becomes true HSR, the travel time would go down to about 4 1/2 hours from downtown KC to downtown CHI, which could compete with air travel.

  • Tom

    I’m sorry but I think this series of articles is making a ton of questionable assumptions. First, 1 out of 4 people in the region riding this thing every year seems high to me. Second, the time it will take to ride the train will increase if the number of passangers increases (security will become tighter, stations will become more crowded, etc). Third, how many private companies are really in a position to buy and operate HSR trains? And how many of them would be willing to do so on a relatively minor route (compared to say, one based in NY or Chicago) before commiting to a bigger route first? Fourth “no cost to the tax payer” doesn’t include paying off the substantial costs of building the rail lines in the first place, which would probably be payed for by issuing multi-year bonds.
    I think HSR is a cool idea, but these projections seem off to me.

    • FrankDeGraaf

      Tom, thanks for the comment. Assumptions are always questionable but as long as we don’t have real numbers that’s all we have.

      The projections are based on Northeast corridor numbers, where no real HSR is offered (ridership would be much higher if there were) and not on true HSR ridership numbers in Europe, Asia, and elsewhere.

      Take the French TGV between Paris and Lyon: Servicing a catchment area of 11 million on a 265-mile line, it attracted 10 million passengers after one year in operation, and 18 million after seven years. Had I used those numbers for the KC-STL catchment area of 5 million it would have resulted in 4 times the ridership I projected for this article.

      Frequencies in Europe and Asia are many times the frequency of the initial projections for KC-STL, without any increase in time to get on the train.

      In part one we discussed some financing options (tax measure / bonds.) Regardless of how to pay for it; building highways and airports is equally costly (the 12-mile renovation of I-64 at $535 million = $ 44 million per mile comes to mind) and will not earn back their investment from user fees.