So, just for fun, I worked out which stations have high platforms and which have low platforms.
On the Empire Corridor, it's high in Metro-North territory (NY Penn through Poughkeepsie), but north of that, the only high platforms are Albany, Schenectady (coming soon), Syracuse, and Rochester (coming soon). In particular, Rhinecliff and Hudson still have low platforms.
On the Adirondack, Montreal has high platforms but everything in between there and Schenectady does not.
On the Keystone, there are *plans* for high platforms everywhere from Harrisburg to Philadelphia, but they aren't *funded*. (Barclur listed the unfunded stations.)
On the NEC itself, the low-level stations are:
Westerly, RI
Mystic, CT
Cornwells Heights, PA
Newark, DE (there are plans for this but they keep changing and being delayed)
Aberdeen, MD (I found a 2012 study for "Aberdeen Station Square")
and of course *some* of the tracks at DC Union.
There seem to be no plans for high-level platforms at Westerly, Mystic, or Cornwells Heights.
The Springfield corridor is all low-level. The commuter rail funding is only going to build high-levels as far north as Berlin. Springfield itself should also get high-levels as part of the endlessly delayed station reconstruction project. (It is hoped that the unfunded highway relocation project in Hartford will involve high-levels there. This still leaves Windsor and Windsor Locks unfunded, though!)
In short, there is a lot of platform work necessary before Amtrak can run all-stops service on trains without traps, on any corridor.
The Palmetto, Meteor, Star, etc are not operating at a surplus or getting close to "breaking even". Some of the eastern LD trains may be covering their direct costs, which is a useful metric when to comes to looking at possible service expansions. But until they cover their fully allocated costs which includes all the overhead, they are losing money.
How can I put this politely?
****.
I apparently understand more about the economics of business lines than you do.
When you talk about an operation "losing money", what do you mean? I believe the actual question is:
"If we discontinued this operation, would the bottom line
(a) improve (i.e. the operation is "losing money" for us)
(b) get worse (i.e. the operation is "making money" for us)
?"
For a train which is covering its avoidable costs, the answer is that discontinuing it would make the bottom line worse -- the train is "making money".
Direct costs are nowhere near a perfect proxy for avoidable costs, but they're the best I've seen. They're certainly a better proxy than "fully allocated costs", which are nonsense.
Larding services up with overhead is bad accounting when you're making business decisions. There are clear case studies of several railroads which made *bad business decisions* based on looking at *profitable* lines which appeared unprofitable after they'd been larded up with overhead. They kept cutting the supposedly "unprofitable" lines, each time they did this, their bottom line got worse.
It is a terrible, terrible business mistake to look at the operations that way. That's why I speak so strongly about it.
Now, do the Star, Meteor, and Palmetto as a *group* "make money"? I'm not actually sure -- since nobody's released direct-costs numbers for fiscal years after 2012 -- but likely not. You could close a lot of stations and thus eliminate some overhead costs if you discontinued *all three* of them. (Though you'd have to relocate Hialeah; so many overhead costs would not go away.) *Individually*, however, the Palmetto and Meteor make money. You can't reasonably count the shared costs towards either one of them individually if you're considering whether to discontinue that one or not.
Here's another way to see why your thinking is simply wrong. Suppose you have three services which which you are claiming "lose money". Suppose you add a fourth service, with the *same* variable costs and the *same* revenues as each of the other three, and now all four of the services "make money". Does this make any sense? No.
What's really going on is that all the services were "making money" all along -- they just weren't making enough money to cover overhead. Expansion of services allowed overhead to be covered. That's the correct way to look at it if you're looking at it as a business matter.
Another way to put it: if you cancelled the entire eastern long-distance network, the cost of the reservations system (which is partly "allocated" to them) would... remain exactly the same and be reallocated to other trains. Whoops.
The only legitimate purpose of overhead allocation is when figuring out how much of a markup you need to charge when doing *contract services*. If you do nothing but contract services, then your markup over avoidable costs on each contract has to include an amount to cover overhead, and to be fair to your customers, you want to spread it evenly over all of them. If you underestimate the number of contracts you're going to get this year, you'll underallocate the overhead and lose money; if you overestimate the number of contracts you're going to get this year, you'll overallocate the overhead and gain money. But it's still a useful tool for this purpose. As such, I'm not going to complain about Amtrak's allocation of overhead to the state-contracted trains; it's an attempt to get the states to cover part of the overhead costs, and it's a negotiated result.
By contrast, if you're running operations on your own account, allocating overhead is silliness and leads to bad business decisions. Please note that Amtrak is running the long-distance trains and the NEC on its own account.
This overhead allocation is the same idiotic thinking which causes people to claim that "Amtrak loses $XXX on every passenger, so if they get more passengers they'll lose more money". It really, really doesn't work that way, and it's misleading to think about it that way.
From a business point of view, what is going on with the best-performing eastern "long distance" trains is that they are profitable to operate, but have very low gross margins. So they do not contribute much to the extremely large overhead of running a railroad. (This seems to be the current problem on the Palmetto, Meteor, and LSL.)
By contrast, a bunch of the western trains seem to be nowhere close to covering their variable costs.
I don't know what the situation was back when David Gunn foolishly cancelled the Three Rivers. It's possible that it was actually losing money, since it seems like all the trains had lower ridership back in 2004. Or maybe it wasn't. It would certainly be making money now. Would the "fully allocated" numbers show that? Depends on what arbitrary allocation procedure was used.
You can make any train look unprofitable by screwing around with overhead allocation. It's not real, it's just confusing. What is real is that some trains have much better gross margins than others. But Amtrak is no longer publishing numbers which allow us to work out the gross margins.