The Antiplanner rarely responds to comments in a post, but Andrew asked many good questions and requested a lot of background information last week. Most of his questions are answered by citations in the report, but since he did not seem to understand those citations, here are my responses.
1. “Intercity buses carry at least 50% more PM than Amtrak in Amtrak’s showcase Northeast Corridor.” “How is this computed and what are the data sources? What trains are you including in the Amtrak total vs. what buses?
I compiled the on-line schedules for what turned out to be sixteen different bus companies for the week of May 15 to 21. The schedules included all buses connecting Northeast Corridor cities: Boston, Providence, New Haven, New York, Newark, Philadelphia, Wilmington, Baltimore, and Washington. I used Google maps to calculate the bus miles between these cities. I then calculated seat miles assuming that premiere lines like Bolt, DC2NY, and Vamoose have 50 seats, Megabuses have 79 seats, and Chinatown buses have 56 seats. My numbers may be an underestimate as some companies may not post their schedules on the web.
To convert to passenger miles, I assumed the seats are 60 percent full. According to the American Bus Association, this is accurate for the major carriers but conservative for the Chinatown buses. Even if the buses are only 50 percent full (which is Amtrak’s average), they carry far more people than Amtrak.
My Amtrak numbers come from page C1 of Amtrak’s 2010 performance report. The September report includes data for the entire fiscal year. Seat miles and passenger miles can both be calculated from this table. I only included Northeast Corridor trains, not trains such as the Crescent, but I also did not include buses such as New York to Raleigh or New York to Atlanta.
2. “60% less energy per PM than Amtrak.” “Again, sources for fuel consumption by the buses and Amtrak? Is this all of Amtrak vs. “all buses? What is “all buses? Source of PM for buses?
This number is based on a 2008 study by M. J. Bradley & Associated commissioned by the American Bus Association. But approximately the same number was used in the 2007 Transportation for Tomorrow report, and it did not cite a source. The number includes all “motorcoaches,” including intercity buses, charter buses, tour buses, and some commuter buses. The American Bus Association has an annual census of motorcoach operators which estimates passenger miles and calculates that a little over a quarter of those passenger miles are scheduled intercity buses.
3. “Intercity buses had been declining since at least 1980.” “Or maybe much earlier? Wikipedia article on Greyhound notes a decline beginning in the 1950s. It’s hard to see how Amtrak “exacerabtes” the decline when both modes were declining in tandem, and Amtrak and Greyhound were trying to cooperate to promote intermodalism.
Yes, the decline probably began earlier, but I didn’t have data to verify it. Greyhound and Trailways still had pretty strong systems in the 1960s, but they basically crashed and burned in the 1980s, and Amtrak had something to do with that, though it was obviously not the only factor. Airline deregulation was probably a bigger factor: per-passenger-mile airfares went from twice bus fares in 1965 to only 15 percent more than bus fares in 1990.
4. “Motorcoach companies pay a lower federal fuel tax of just 7.3 cents per gallon.” This is a huge subsidy, considering the axle load of buses (and thus their damage to pavements and bridges) is equivalent to heavy trucks. Trucks are taxed at 24.4 cents per gallon. 17.1 cents per gallon difference @ annual consumption of 1 billion gallons claimed by BTS is $170 million.
This subsidy is included in the Nathan & Associations report on transportation subsidies. That report found total motorcoach subsidies to be a little more than $60 million a year, a little more than a quarter of which can be attributed to scheduled intercity buses.
5. “Many carriers provide more buses Friday through Sunday than Monday through Thursday.” “Showing empirically what we all know anecdotally that buses are used by cheapskates out for a weekend and not by regular business and leisure travelers, and that their appeal is mainly on ticket price because they cannot compete on speed.
Another way of putting this is that people will ride a train when their employer is paying for it but take a low-cost bus when they have to pay for it themselves. Even another way of putting this is that the capital costs of buses are low enough that they can flexibly respond to day-to-day differences in demand.
6. “ABA report that they fill an average of 60% of their seats.” What is this statistic? Do they sell 60% of seats on a trip with multiple stops, or do they fill 60% of seat miles with a passenger riding a mile? Any bus with multiple stops will have a different load factor than a ticket/seat ratio, because seats will not be used the entire length of the journey.
This means that total passenger miles will be 60 percent of total seat miles.
7. “By comparison, Amtrak reports that it offered” These numbers only include trains on the NEC trunk and do not include additional service provided by trains continuing to Springfield, Harrisburg and Richmond.
I also did not include buses to Springfield, Harrisburg, and Richmond.
8. “Amtrak carries little more than two-thirds as many passengers as intercity buses.” How certain are we of that statistic? Greyhound’s nationwide ridership is around 22 million, and Megabus is just 4 million per year each for all routes, per their annual report.
My calculations are passenger miles, not ridership. In the Northeast Corridor, the average length of bus trips appears to be about the same as the average length of train trips, so if buses carry more passenger miles than Amtrak, they also carry more passengers.
My Northeast corridor company-by-company seat-miles in millions are: Greyhound: 679; Megabus: 670; Bolt Bus: 564; Peter Pan: 546; Apex: 344; New Century: 329; Fung Wah: 181; Eastern: 163; Lucky Star: 143; AA: 96; Vamoose: 70; Washington Deluxe: 53; MVP: 53; Tripper: 42; DC2NY: 34; LimoLiner: 17.
9. Are you including gambling oriented buses to Foxwoods, Mohegan Sun, and Atlantic City?
No, only scheduled intercity buses between Boston, Providence, New Haven, New York, Newark, Philadelphia, Wilmington, Baltimore, and Washington (and suburbs of Washington and New York).
10. “Because the new model of bus service is mostly based on nonstop buses.” In other words, it is not directly comparable to Amtrak because the buses only offer service to major cities and skip over intermediate points and intermediate markets.
Actually, it is superior to Amtrak because it provides non-stop service while Amtrak has to stop. That’s one of the advantages of buses over trains: because motorcoaches are smaller, they can offer more precise destinations.
11. “Amtrak requires 5 [Andrew means 6] hours and 30 minutes to carry passengers from Chicago to Detroit.” The Amtrak schedule clearly shows 5 hours 30-38 minutes to downtown Detroit. You are deliberately distorting that by using suburban Pontiac times as the Detroit time.
You are right; Amtrak’s Chicago-Detroit trains take 5 hours and 30 minutes, not 6 hours and 30 minutes as my report says in one spot.
12. “Reduce this travel time by 12 minutes” We went over this when you posted the article on Michigan HSR. The stated intention of MIDOT is to reduce running times by 50-60 minutes to 4 hours 39 minutes on this corridor, and almost all aspects of the plan have been fully funded. Its clearly shown on Page 10 of their Corridor Service Development Plan in the proposed schedules.
We’ll see what happens when the new timetables come out.
13. “Both regional trains and buses take from 4 hours” The buses run nonstop, the trains make 12 stops and have the same running time and thus serve many more markets and obviously can get you closer to your destination.
See my response to 10 above.
14. “New model fares average about 7 to 10 cents per PM.” Except when they don’t on the luxury type services where they are up to 40 cents per PM, as on LimoLiner.
The first-class operators–LimoLiner and Vamoose Gold–have 0.6 percent share of the market, so their fares are not going to change the average all that much. Megabus and Bolt probably sell more $1 tickets than Vamoose Gold’s and LimoLiner’s combined ridership.
15. “Amtrak fares per PM” Which is distorted by including high fare Amtrak routes like Acela, most NEC Regional, and Auto Train with the general run of the mill Amtrak routes. Most Amtrak routes are 13 to 17 cents per PM in revenue.
Amtrak fares average 42 cents a passenger mile for the “regional trains” in the Northeast Corridor and 74 cents a passenger mile for the Acela. Other corridor trains average 20 cents a passenger mile. Only the long-distance trains average as low as 16 cents a passenger mile.
17. “Average cost of 15 cents per PM” I.e., the same as the typical Amtrak fare noted above except specialty/first class/high speed type services able to charge more money.
As noted above, Amtrak fares are higher than 15 cents for all but some long-distance trains and a handful of corridor trains. Also, we have to distinguish between “fares” and “cost”: the cost of driving, including highway subsidies, is 15 cents a passenger mile, while Amtrak costs are the 28-cent-per-passenger-mile fare plus the subsidy which itself is more than 25 cents per passenger mile.
18. “It probably would offer more non-stops.” What is stopping them? Next to no cost of entry, cheaper fares than Amtrak, faster running times. Shouldn’t they just eat the market up? Or maybe there is very little demand for buses?
What is stopping them? How about five heavily subsidized Amtrak trains a day?
19. “Similarly, the withdrawal of Megabus from the California market after less than a year … is partly due to … subsidies to short-distance corridor trains …†Except that Megabus was focused on routes like LA-Phoenix, LA-Las Vegas, LA-San Diego, and LA-Bay Area. Amtrak only offers service on one of those routes – LA-San Diego. How could subsidies for the San Joaquins or the Santa Barbara service cause a problem for Megabus routes to Las Vegas and San Francisco?
Note that Megabus’s successful Midwest and Northeast routes are all around 200 miles, while the California routes Andrew mentions were much longer. If Amtrak had not previously captured the short-distance routes (LA-Bakersfield, Sacramento-Fresno, etc.), then Megabus would probably have done much better.
20. “Offer little more convenience than a relatively unsubsidized bus system.†You mean, relatively little more convenience aside from serving intermediate points and providing travel times half those of the bus and providing food and a nicer on-board environment? Nothing major there, right?
If I am going from New York to Washington, the fact that the train stops at intermediate points is an inconvenience, not a feature. Yes, trains have a nicer on-board environment–at a much higher cost. I am not sure if what Amtrak serves in its corridor trains deserves to be called “food,” but if it does, you should compare Amtrak with luxury buses like LimoLiner, which is $10 less than Amtrak in the Boston-to-New York market, has a great on-board environment, and the food it serves (which appears to be roughly equal to Amtrak’s food) is included in the fare.
21. “The average Amtrak intercity train … emitted 3.2 times as much CO2.” The average Amtrak intercity train runs off electric power, because most of its trains are in the NEC. The electric supply of the NEC is mainly Susquehanna River hydropower and Exelon Nuclear plants. So no CO2 emissions. But why bother with such facts?
States in the Northeast Corridor get half or more of their electricity burning fossil fuels. Wherever Amtrak gets its electricity, it is displacing other customers in the region who buy fossil-fuel-generated electricity. But why bother worrying about opportunity costs?
22. “Passenger rail lines require so much infrastructure, while highways — the basic infrastructure for buses — are shared with cars and trucks.” Other than a handful of commuter lines, every passenger rail line in the US is shared with freight trains, including the Northeast Corridor. The infrastructure requirements of most passenger rail lines are marginal because the tracks would be there anyway for freight.
Actually, they are far from marginal because rail lines must be maintained to higher standards to provide passenger services at any reasonable rate of speed (most freight trains only go 40 to 45 mph). In addition, many of the most popular passenger rail routes are also being used to capacity or near capacity by freight trains, which is why the main object of the Illinois “high-speed rail” project is to increase the route’s capacity, not to increase speeds. Freight service in the Northeast Corridor, by the way, is pretty marginal: generally no more than two through trains per day and possibly some locals.
23. “Amtrak fills only about half its sets.”Because Amtrak makes intermediate stops which non-stop bus companies ignore, so the peak loading point is not the end points, so seats inevitably are empty more of the time because they are being made available for non-endpoint trips. By serving markets other than only the biggest cities, Amtrak has to run more equipment. Like Greyhound too.
It sounds like you are saying that the rail model is defective because it must make lots of intermediate stops, which both slow it down and reduce occupancy rates. Since rail’s capital costs are also much higher, buses have multiple advantages over rail.
24. “Allowing companies to use any curbside” In a normal city downtown, the value of curbside is established in part by parking meters which charge $1-3 per hour per spot. The new model bus companies claim the curbside as their own for free for no payments. A 500 foot block should have 20+ parking spots and a theoretical value of over $500 per day. A single bus company with 10 curbside blocks at 10 destinations dedicated to them is getting a $2 million annual subsidy.
Which is why I recommended charging market rates for curb rights.
25. “Cities to lease curb rights.” Will parking companies also get to bid for these spots? How about nearby retail businesses or restaurants?
Why not? I agree with Don Shoup: the government should not provide parking at below-market rates.
26. “Enforcement of existing safety rules” When is that going to happen? How about a speeding and operating rules fine regime similar to what FRA imposes on Amtrak engineers, including heavy personal liabilities for infractions, week long annual rules reviews, etc.? Has anyone every seen a bus pulled over for a speeding ticket?
It is already happening.
27. “Between 1999 and 2008, Amtrak passengers suffered 81 passenger fatalities”Or maybe not? Where were all of these wrecks and dead passengers? Amtrak’s total fatality count for its entire 40 year existence is barely over 100. Most of those were at Bayou Canot in 1993, and much of the remainder at Chase in 1987 and Salem in 1971.
My data came from “passenger fatalities” shown in this table of “Railroad Passenger Safety Data.” Since commuter-rail and other rail transit fatalities are separately shown in this table, I presumed–perhaps erroneously–that the first table is all Amtrak.
28. “Seat belts might more effectively be applied to intercity trains.” Well, that could be the case if the death rates were what you state, which they aren’t, and it could be shown that the deaths on trains resulted from being thrown from your seat, and not from crushing or burning or smoke inhalation, which appear to be the actual causes in most train wrecks.
Yes, but the same is true for buses: most of the bus fatalities were not scheduled intercity buses and many were not from people being thrown from their seats.
29. “Alternatives for low-income people.” How about alternatives for the middle class and upper income people? Most people are not low-income.
Some people argue that low-income people deserve subsidies to give them a leg up, but no one argues that middle- and upper-income people deserve subsidies just because of their incomes. The question is: why are we heavily subsidizing trains and then charging fares that make them accessible only to middle- and high-income people, while shutting buses out of many markets even though the fares they would charge are much lower?
30. Conclusion. With essentially no cost of entry to start service and supposed great efficiencies everywhere, there is very little standing in the way of your buses taking over the world vision except for the aversion of most people to actually riding them due to slow speeds, tiny seats, and the existing socio-economic background of the clientele.
The most profitable businesses serve a broad market: GM sells both Chevrolets and Cadillacs; Toyota sells Scions and Lexuses. With Amtrak capturing the middle- and high-end of the market, buses have no opportunity to do anything but serve the low end of the market.
Private operators such as the Hampton Jitney show that buses can successfully attract the high end of the market. Yet with heavily subsidized Amtrak competing against them in most major corridors, bus operators have to settle for thedregs. No wonder buses have such a bad reputation, and no wonder their operating margins are so slim that companies like Megabus don’t serve California, the Pacific Northwest, and other major corridors.
Quoting The Antiplanner quoting Andrew:
“Motorcoach companies pay a lower federal fuel tax of just 7.3 cents per gallon.â€
This is a huge subsidy, considering the axle load of buses (and thus their damage to pavements and bridges) is equivalent to heavy trucks. Trucks are taxed at 24.4 cents per gallon. 17.1 cents per gallon difference @ annual consumption of 1 billion gallons claimed by BTS is $170 million.
When traveling by highway from Washington, D.C. to New York, N.Y., much of the trip is on roads that do not rely on motor fuel taxes for ongoing maintenance and operations, for they are toll roads.
Based on what the various toll agencies publish online, I estimate that a northbound bus trip pays about $42.75, and a southbound bus trip is charged $38.75 (difference in tolls is because some of the toll roads and toll crossings only charge a toll in one direction).
I assumed that buses use the following toll crossings and toll roads:
– Baltimore, Maryland – Harbor Tunnel or Fort McHenry Tunnel
– Maryland’s JFK Highway (I-95 between Baltimore and the Delaware line)
– Delaware Turnpike (I-95 from the Maryland line to I-295)
– Delaware Memorial Bridge (I-295 between Delaware and New Jersey)
– New Jersey Turnpike (from Exit 1 (Delaware Memorial Bridge) to Exit 16E (Lincoln Tunnel))
Tolls are addition to, not in place of, federal and state motor fuel tax on the Diesel fuel that they consume.
So the subsidy is not quite as high as Andrew asserts.
Compare and contrast with Amtrak, where federal taxpayers, and not Amtrak customers, pay to repair and overhaul the Northeast Corridor tracks and overhead catenary on which Amtrak runs its “profitable” trains.
CPZ:
When traveling by highway from Washington, D.C. to New York, N.Y., much of the trip is on roads that do not rely on motor fuel taxes for ongoing maintenance and operations, for they are toll roads.
That doesn’t make any difference to those of us in cars, who must pay both the toll and the full motor fuel tax. Why should it matter for buses? If the tolls didn’t exist, motor fuel taxes would obviously need to be much higher.
Based on what the various toll agencies publish online, I estimate that a northbound bus trip pays about $42.75, and a southbound bus trip is charged $38.75 (difference in tolls is because some of the toll roads and toll crossings only charge a toll in one direction).
Compared to a car or trucks toll, the bus is again getting subsidized:
Car
$8.00 Lincoln Tunnel (NB only)
$8.20 NJ Turnpike
$4.00 Del Mem Bridge (SB only)
$4.00 Delaware Turnpike
$5.00 Maryland Turnpike (NB only)
$2.00 Baltimore Tunnels
Round trip – $45.40
Bus
$6.00 Lincoln Tunnel (NB only)
$12.65 NJ Turnpike
$15.00 Del Mem Bridge (SB only)
$6.00 Delaware Turnpike
$15.00 Maryland Turnpike (NB only)
$6.00 Baltimore Tunnels
Round trip – $85.30
3 Axle Truck
$24.00 Lincoln Tunnel (NB only)
$23.10 NJ Turnpike
$15.00 Del Mem Bridge (SB only)
$6.00 Delaware Turnpike
$15.00 Maryland Turnpike (NB only)
$6.00 Baltimore Tunnels
Round Trip – $124.20
Buses weigh about 10 times the weight of the average passenger vehicle, but pay less than twice the tolls. They pay just 2/3 of the tolls of a similar size truck. No matter which way you slice it, its more free rides for the bus.
On the PA Turnpike, buses pay three times the toll rate of cars and on the New York Thruway, its four time the rate.
Tolls are addition to, not in place of, federal and state motor fuel tax on the Diesel fuel that they consume. So the subsidy is not quite as high as Andrew asserts.
Every other road user pays more tolls than the bus on a weight basis, so the bus is actually getting more subsidies. They also pay more gas taxes, and no one gets a discount on gas taxes for driving toll roads.
federal taxpayers, and not Amtrak customers, pay to repair and overhaul the Northeast Corridor tracks and overhead catenary on which Amtrak runs its “profitable†trains.
A lot of NEC maintenance costs are paid from operating income of the Amtrak trains, charges to Amtrak long distance trains using the NEC (16 per day), real estate rents and air-rights in and around stations, and freight and commuter rail access fees (which total over $100M). Most of the latter items are never considered in the discussion, of course.
Randall:
Thank you for taking the time to make responses. Here are a few more comments for you to consider:
only included Northeast Corridor trains, not trains such as the Crescent, but I also did not include buses such as New York to Raleigh or New York to Atlanta.
You included buses serving suburban Washington and Philadelphia, but not Amtrak trains to Alexandria/Woodbridge/Quantico/Burke Center/Manassas south/west of Washignton and to Ardmore/Paoli west of Philadelphia. I know it would be difficult to tease out the results without access to better Amtrak numbers, but it bears noting it is not an apples to apples comparison. Amtrak got 206,000 ons+offs at Ardmore and Paoli and 192,000 from the suburban Washington stops. Obviously some of that traffic is not heading towards NY/DC, but some is.
Another way of putting this is that people will ride a train when their employer is paying for it but take a low-cost bus when they have to pay for it themselves. Even another way of putting this is that the capital costs of buses are low enough that they can flexibly respond to day-to-day differences in demand.
Amtrak responds to the weekend demand by charging higher prices to drive off marginal traffic (and the trains still get higher ridership). If Amtrak responded by adding cars and keeping prices down, you’d probably see less bus peak traffic.
My Northeast corridor company-by-company seat-miles in millions are: Greyhound: 679; Megabus: 670; Bolt Bus: 564; Peter Pan: 546; Apex: 344; New Century: 329; Fung Wah: 181; Eastern: 163; Lucky Star: 143; AA: 96; Vamoose: 70; Washington Deluxe: 53; MVP: 53; Tripper: 42; DC2NY: 34; LimoLiner: 17.
Are you sure the Greyhound and Peter Pan numbers are not being double counted? If you look at their schedules, you can quickly see many of the trips between Boston, NY, and DC are joint schedules. Also, some Megabus route schedules are interlined.
I also think you are overestimating total ridership from the Seat Miles. As I noted, Greyhound’s annual report gives nationwide ridership of 22 million. Megabus was up to around 3.9 million per year total for the US AND Canada (they reported in July 2011 carrying their 10 millionth passenger in the US AND Canada since 2006 and their 6 millionth in July 2010). Wikipedia reports Peter Pan’s total ridership is “over” 4 million. Looking at your seat mile totals for the NEC, its very difficult to see how you run up 15 million annual riders from the big three and the lesser companies when it is obvious that the lesser companies are very much lesser. I understand your math, but your totals do not jive with the overall copany ridership being reported. If you assume, for example, that 1.5 million of Megabus’ 4 million riders are on the NEC stem, and then project total ridership for your 4 billion seat miles from that, you get just 9 million riders. Your mileage may vary with other assumptions.
Actually, it is superior to Amtrak because it provides non-stop service while Amtrak has to stop. That’s one of the advantages of buses over trains: because motorcoaches are smaller, they can offer more precise destinations.
Offering fewer stops doesn’t seem like an advantage to me. If I live near BWI, why would I want to go to Baltimore to get transportation?
What is stopping them? How about five heavily subsidized Amtrak trains a day?
But as you note, the bus is faster, cheaper, and more frequent. In theory, it should eat up the market. The problem is that as Megabus notes in its annual report, 92% of its riders are choosing it based on cost. Which means to the non-cost discretionary traveller, Megabus is simply not considered an option.
States in the Northeast Corridor get half or more of their electricity burning fossil fuels. Wherever Amtrak gets its electricity, it is displacing other customers in the region who buy fossil-fuel-generated electricity. But why bother worrying about opportunity costs?
Amtrak sources most of its electrical power specifically from dedicated turbines at Safe Harbor Dam (40% of total use and soon to be owned by PECO) and from feeds from PECO/Exelon around Philadelphia (45% of total use), which is an 80%+ nuclear/hydro utility. The PECO/Exelon feed from its nuke/hydro network replaced obsolete dedicated coal/oil power stations which are now either peak load only or retired. There isn’t an opportunity cost as it was actually the electric load from the PRR and the trolley companies that permitted the creation of a widespread electrical network that could be used by others – the PRR railroad transmission network was one of the first in the country. While there may be a lot of fossil fuel use in the northeast, it is by utilities like Pennsylvania Power and Light, not PECO, the utility providing the bulk of Amtrak’s power and serving Amtrak and its portion of the region.
Actually, they are far from marginal because rail lines must be maintained to higher standards to provide passenger services at any reasonable rate of speed (most freight trains only go 40 to 45 mph).
Nationwide, typical over the road freight speeds are 70 mph intermodal, 60 mph manifest freight, and 50 mph mineral/grain trains.
Any freight service over 40 mph and up to 60 mph requires maintenance of track to FRA Class IV standards, which permit 60-80 mph passenger service. 60-70 mph intermodal trains require maintenance to Class V standards, which permit 80-90 mph passenger service. There is no higher maintenance standards in that sense from running passenger trains.
The little bit of extra maintenance comes when a curve is given more superelevation than needed to accomodate higher passenger speeds. The higher superelevation causes accelerated rail and tie wear in the curves.
In addition, many of the most popular passenger rail routes are also being used to capacity or near capacity by freight trains
This is mostly a myth. The normal variation in freight traffic over a week can be as much as 20 trains per day on a line that might peak at 60 trains per day. Obviously a couple of trains per day is not going to push this line over the edge. Where there are problems are a few places like Chicago to Porter, IN, where there are 14 Amtrak trains and 100+ freights on a line that has been over capacity for decades. Here the freight railroads have decided to shoot themselves in the foot in the hopes of getting public money out of the chaos instead of having to spend their own money.
Freight service in the Northeast Corridor, by the way, is pretty marginal: generally no more than two through trains per day and possibly some locals.
It depends on the section. There are about 12 freight trains per day between Baltimore and Wilmington heading to Perryville and then up to Harrisburg.
It sounds like you are saying that the rail model is defective because it must make lots of intermediate stops, which both slow it down and reduce occupancy rates. Since rail’s capital costs are also much higher, buses have multiple advantages over rail.
No, the whole point of passenger rail is to make the intermediate stops that buses and planes do not want to make because they could never be economical on a point-to-point network. The capitol costs are only higher if you insist on a dedicated network that cannot share tracks with freight – like insisting on dedicated busways, and you force the rail network to be so small that it cannot sustain a supplier base that provides products at reasonable prices. Otherwise, the tracks are there waiting to be used.
Some people argue that low-income people deserve subsidies to give them a leg up, but no one argues that middle- and upper-income people deserve subsidies just because of their incomes. The question is: why are we heavily subsidizing trains and then charging fares that make them accessible only to middle- and high-income people,
The argument for rail subsidies is not an income based argument, but to provide an alternate means of transportation apart from the road and air networks and their congestion. There are plenty of poorer people on the train too once you stop looking only at the NEC. The train actually successfully serves all socio-economic levels.
while shutting buses out of many markets even though the fares they would charge are much lower?
How are buses being shut out of the market? Anyone can start running buses if they want, and you note their inherent economic advantages. The problem is a lot of competition and little profitability for a limited market.
Private operators such as the Hampton Jitney show that buses can successfully attract the high end of the market. Yet with heavily subsidized Amtrak competing against them in most major corridors, bus operators have to settle for thedregs.
The high end bus market only exists in New York City where there are wealthy people without cars. It is nonexistant everywhere else. The settling for dregs is not the result of Amtrak, which as you note outside the NEC has higher fares and really serves a market of small towns the buses like to skip, but of the inherent nature of the product.